[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7B48M35DPRDCEWGRBGYCTQJ":3,"skill-same:01M7B48M35DPRDCEWGRBGYCTQJ":46},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7B48M35DPRDCEWGRBGYCTQJ","verbs",null,"Every nemia verb you run, by who answers for it — a read, yours unasked, one that prompts the person, one that answers what waits on them — with its usage and help.","emyann\u002Fnemia","emyann","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F4748419?v=4","01M7B458J23KJ5S5AEBR86G7RN","Yann RENAUDIN","01M7B458JPM0CCBDP4KCR6NATT",3,"MIT","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Fagent-orchestrator\u002Fnemia\u002Fskills\u002Fverbs\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: verbs\ndescription: Every nemia verb you run, by who answers for it — a read, yours unasked, one that prompts the person, one that answers what waits on them — with its usage and help.\n---\n# The verbs\n\nEvery verb `nemia` serves you, by who answers for it, with its usage and help as `nemia --help` prints them. `nemia \u003Cverb> --help`\nsays more of one that has a manual. With a shell, run `nemia \u003Cverb>`; on the MCP port, call the tool its `on the port` line names,\nwhose schema says its arguments. A verb said `CLI only` runs at a shell alone, for the reason given.\n\nGenerated from the command table, never edited by hand: from the repository's root, `node -e \"import('.\u002Fpackages\u002Fagent-orchestrator\u002Fsrc\u002Fnemia\u002Fverbs.verb.ts').then((m) => process.stdout.write(m.verbsSkill()))\" > packages\u002Fagent-orchestrator\u002Fnemia\u002Fskills\u002Fverbs\u002FSKILL.md`.\n\n## Reads\n\nThey write nothing: run them unasked, as often as you need. One that sets with some words prompts the person given them.\n\n```text\n  nemia concluded \u003CrunId>\n                               what a run started with --from \u003CrunId> would be handed of it, now\n                               on the port: concluded\n  nemia inbox\n                               the cards still in play: name, first line, workflow when set,\n                               and started as \u003CrunId>, \u003Cphase> (nemia inbox --help); --all: those whose\n                               run ended too; -l: the cards whose labels match; --show-labels, -L \u003Ckey>:\n                               their labels on each line; -o wide: every column\n                               on the port: inbox, inbox_card\n  nemia factory\n                               the store's factories: the default first — every run and card that\n                               names no factory, read, never written — then each by name with its\n                               repositories (nemia factory --help)\n                               on the port: factory\n  nemia factory ls\n                               the same, by its name\n                               on the port: factory_ls\n  nemia factory diff \u003Cname> --to \u003Cx.y.z> [--template \u003Cname|folder|archive>] [--set \u003Ckey>=\u003Cvalue> ...] | \u003Cname> --template \u003Cname|folder|archive> [--set \u003Ckey>=\u003Cvalue> ...]\n                               preview a templated factory's upgrade: what merges cleanly, and each\n                               conflict with its sides; onto a template of another name, a switch\n                               with no common base; or a hand-made factory's takeover by a\n                               template, and the plan nemia factory adopt takes; nothing is written.\n                               To change a parameter alone: --to the version applied, and\n                               --set \u003Ckey>=\u003Cvalue> for each; it prints the line that files it\n                               on the port: factory_diff\n  nemia factory servers \u003Cfactory>\n                               a factory's own MCP servers, each with the state its harness reports\n                               on the port: factory_servers\n  nemia factory rules \u003Cname> [--file \u003Cpath>|- | --clear]\n                               a factory's rules and policies: words told to every agent hired into\n                               it, after its custom instructions and before the step's own note,\n                               kept by the store with the factory. Bare, prints them; --file writes\n                               them (- for stdin); --clear removes them. Words only: a frontmatter\n                               block is refused. A run keeps the rules it was started with\n                               reads bare; prompts the person given --file, --clear\n                               on the port: factory_rules, factory_rules_put, factory_rules_delete\n  nemia factory instruction \u003Cname> [--path \u003Cpath> | --store | --clear]\n                               where a factory's instruction is kept, what Nemia is to know of the\n                               factory (its agents are told its rules: nemia factory rules): bare,\n                               prints it; --path reads it from a file in the factory's first folder,\n                               --store keeps it in the store, --clear reads NEMIA.md in the first\n                               folder again\n                               reads bare; prompts the person given --path, --store, --clear\n                               on the port: factory_instruction, factory_instruction_put, factory_instruction_delete\n  nemia template lint \u003Cfolder|archive>\n                               check a template's folder or archive: every problem at once, each at its\n                               file and line\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia ls [\u003CrunId>] [-l \u003Ckey>=\u003Cvalue>] [--show-labels] [-L \u003Ckey>] [-o wide]\n                               runs and their steps — a run waiting on another says waiting on \u003Crun>,\n                               one the cap on runs in flight holds waiting for a run slot (2 of 2 in\n                               flight); what each step's tries and the run have cost, where measured;\n                               last, whether a nemia up ticks the store and when its last pass was;\n                               first, while the store is paused, Pausing: N finishing and each\n                               session, check or landing at work, or Paused once none is, then each\n                               check it holds, a landing with its merge staged in the repository's\n                               own working tree (MERGE_HEAD) among them; a step whose check it holds\n                               says waiting on the pause. -l: the runs whose labels match (team=core,\n                               given again for another); --show-labels: each run with its labels;\n                               -L \u003Ckey>: that label beside each run, given again for another; -o\n                               wide: every column, each step among them\n                               on the port: ls, ls_run\n  nemia show \u003CrunId> [--raw]\n                               the run as one object (nemia\u002Fv1 Run); --raw: its record, and\n                               what each step was told\n                               on the port: show\n  nemia closeout \u003CrunId> \u003Cstep> [--try \u003Cn>]\n                               what a step's try signed off with: its latest, or --try \u003Cn>\n                               on the port: closeout\n  nemia logs \u003CrunId> \u003Cstep>\n                               that step's session output\n                               on the port: logs\n  nemia stats [--from \u003Cday>] [--to \u003Cday>] [--by agent|step|factory|workflow|repo] [--agent|--step|--factory|--workflow|--repo \u003Cv>] [-l \u003Ckey>=\u003Cvalue>] [--line|--json]\n                               what every run in the store adds up to, by UTC day: rounds to\n                               acceptance of each loop: step, landings of the workspace: base step\n                               and on which try, cost and time per landed run, and each step's\n                               tries, time, cost, checks and gate waits; first the last 14 days\n                               against the 14 before. A number nothing measured reads —, never 0.\n                               --line: one line, for a recap; --json: the whole answer; -l: the\n                               runs whose labels it takes, as nemia ls -l takes them\n                               on the port: stats\n  nemia passes [--last \u003Cn>] [--json]\n                               what the recent passes of the tick cost: their median and slowest\n                               time, a histogram of their times, where the time went phase by\n                               phase, and what the last saw — the runs in the store, its journal\n                               reads, the sessions queued and admitted. --last \u003Cn>: the last n\n                               passes only; --json: the whole answer\n                               on the port: passes\n  nemia series [--range 24h|7d|30d|all] [--json]\n                               the system's series over a range: a pass's time and its phases, its\n                               journal reads, the runs in the store, the sessions queued and admitted\n                               and the longest wait, the load, and the quota used. --range: 24h (a\n                               point a minute, the default), 7d or 30d (an hour), all (a day);\n                               --json: the whole answer\n                               on the port: series\n  nemia agents [--repo .]\n                               the agents a step may name, the repository's then yours: what each is\n                               for, what it says it can do (an entry for each skill, read off its\n                               SKILL.md), and the skills it is given (its skills:)\n                               on the port: agents\n  nemia agents card \u003Cname> [--repo .] [--from repo|yours]\n                               that agent in the shape of an A2A Agent Card, JSON on stdout and\n                               nothing else — every field the protocol marks REQUIRED but\n                               supportedInterfaces, which is left out because nothing here serves an\n                               agent over A2A. The card describes; it is not an address\n                               on the port: agents_card\n  nemia agents sources [--path \u003Cfolder>]\n                               the agent files the store knows of, to copy into your library:\n                               each repository's, yours, and those of a --path \u003Cfolder>\n                               on the port: agents_sources\n  nemia agents custom \u003Cname> [--file \u003Cpath>|-] [--clear] [--repo .]\n                               your custom instructions for that agent: words you add to it without\n                               rewriting it, kept in \u003Cstore>\u002Fcustom-instructions\u002F\u003Cname>.md and told\n                               after its instructions. Bare, prints them; --file writes them (- for\n                               stdin); --clear removes them. Words only: a frontmatter block is\n                               refused. A run keeps the words it was started with; nemia, the\n                               store's own agent, is told them in her open session\n                               reads bare; prompts the person given --file, --clear\n                               on the port: agents_custom, agents_custom_put, agents_custom_delete, nemia_instructions, nemia_instructions_put, nemia_instructions_delete\n  nemia wait [-l \u003Ckey>=\u003Cvalue>]... [--from \u003Ccursor>] [--for \u003Cduration>]\n                               wait for the feed's next line -l takes, and print it with the\n                               cursor after it: from --from (a cursor a wait or nemia watch printed),\n                               else from the feed's end now. It ends on the first lines -l takes, or\n                               after --for (4m by default, at most 270s) with none, and prints the\n                               cursor either way: wait again from it and no line is lost.\n                               -l: as nemia watch -l takes it, given again for another; none takes\n                               every line.\n                               on the port: wait\n  nemia level [manual|semi-auto|auto] [--factory \u003Cname> [--for \u003Cduration> | --end]]\n                               Nemia, the store's own agent: bare, where she is, each factory's level and\n                               how to reach her; a level sets the store's, which every factory follows\n                               unless it has its own; --factory sets that factory's, --for raises it for\n                               a stretch (90m, 2h, 1d; up to 7d) and --end ends the raise. semi-auto keeps\n                               her session: what waits on you (a gate, a step's question, a loop out of\n                               rounds, a cap) is asked of you, and a write that cannot be undone or\n                               changes your settings or a factory; the rest is hers. auto is semi-auto,\n                               but an answer your confirmed list allows in the run's state is given\n                               unasked (nemia policy \u003Cagent>), in a session of hers filed since auto came:\n                               one kept from before asks every answer until filed again (manual, then\n                               auto once it is off). A factory at manual is left alone, and her session\n                               stops once none is above manual\n                               reads bare; prompts the person given manual, semi-auto, auto, --factory, --for, --end\n                               on the port: level, level_set\n  nemia memory [\u003Cfact>]\n                               what Nemia keeps of the store: bare, her facts' names; given one, its\n                               words. Read only: her session alone writes them\n                               on the port: memory, memory_fact\n  nemia policy \u003Cagent>\n                               the agent's rules at auto: the answers it may give unasked, in force once\n                               you confirm them, and what it proposed from your custom instruction,\n                               with the words the machinery cannot hold, kept as guidance\n                               on the port: policy\n  nemia policy reach \u003Cagent> (--factory \u003Cname> | --run \u003CrunId>)\n                               which steps at their gate the agent's rules in force could ever approve\n                               unasked, in a factory or a run: why each other rule cannot (a step approved\n                               before it runs has done nothing to recommend or check; one with no check\n                               never passed one), what keeps a rule from a run now, and the rules that would\n                               on the port: policy_reach\n  nemia service status [--name \u003Cname>]\n                               the login service and the lease: whether it is installed and runs,\n                               the checkout it runs, who holds the store, and lingering\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service addresses [--json]\n                               this machine's addresses, loopback first, one a line: the address, its\n                               interface and its kind (loopback, private, shared, link-local,\n                               unique-local, public), so --bind takes a private one; --json prints\n                               {\"addresses\":[{\"address\",\"interface\",\"family\",\"kind\"}]}\n                               CLI only: it reads the network interfaces of the machine the command runs on\n  nemia doctor [--json]\n                               every piece this server needs before a run: each harness installed and\n                               signed in, the service and its linger, the bind, the machine, the git\n                               identity, each factory's workflows' and repository's needs, a factory, a\n                               newer release — each in place or missing, with the command or fix that\n                               puts it there. --json: the whole read\n                               on the port: doctor\n  nemia readiness\n                               retired: nemia doctor reads what it read\n                               CLI only: retired: nemia doctor is the read, served on the port\n  nemia token ls\n                               the tokens you minted, by name: who holds each, the factory and the\n                               narrowing of an agent’s, and when it was minted. Never a token itself\n                               on the port: token_ls\n  nemia pairings ls\n                               the pairings the store keeps, newest first: the client that asked, and\n                               whether it waits for your approval, is approved, was claimed or lapsed.\n                               Never a code: approve the one the client shows with nemia pairings approve\n                               on the port: pairings_ls\n  nemia claude-token status\n                               whether your subscription token is kept, its file and its mode, when it\n                               was set and the days left of its year, warned of a month before it ends:\n                               never the token\n                               CLI only: the token is the person's own, kept on the machine the store is on and read there from the terminal; no caller of the port reads or writes it\n  nemia version [--json]\n                               this nemia's release, an install's tag or a checkout, and its folder;\n                               the release the store's releases\u002Fcurrent names; and the release the\n                               daemon runs, from its lease, and whether it promises to keep a running\n                               session on its files; and the sessions at work by release and harness\n                               version, off their records; and a newer release, where the daemon's\n                               release check read one. --json: the whole answer\n                               CLI only: it says which code the command runs from; the api answers its own at GET \u002Fapi\u002Fversion\n  nemia admit [--at-least \u003Clevel>] [--auto \u003Cagent>]\n                               a call's admission: the call a session is about to make, on stdin, answered\n                               ask where it answers what waits on you;\n                               --at-least: deny a call acting on a run or a factory below the level;\n                               --auto: let through, for a run of a factory at auto, an answer the agent's\n                               rules you confirmed allow in the run as it stands (nemia policy \u003Cagent>)\n                               CLI only: a session's hook asks it before each call, on its own machine\n  nemia decide \u003CrunId> [--sensitive \u003Cglobs>]\n                               what the run produced, and the choice it leaves you (nemia decide --help)\n                               on the port: decide\n```\n\n## Yours\n\nThey write, and are yours to run unasked, on your judgement: say what you did.\n\n```text\n  nemia start \u003Cbrief.md|card|\"text\"> --workflows \u003Cname> [--repo .] [--no-worktree] [--after \u003CrunId> ...] [--from \u003CrunId> ...] [--input \u003Cfile> ...] [--card \u003Cname> ...] [--gate \u003Cstep> ...] [--model \u003Cstep>=\u003Cmodel> ...] [--factory \u003Cname>] [--brief \"\u003Cbrief>\"]\n                               start a run of a workflow from a brief, a text, or a card of the inbox\n                               --workflows \u003Cname>: the workflow, by its name, as its factory offers\n                               it (nemia workflows lists them); a card's own workflow: where it names\n                               one. A path to a workflow folder, and a folder of cards, are refused:\n                               file each card (nemia inbox add \u003Ccard>) and start it by its name\n                               --after \u003CrunId>: file nothing of this run until nemia decide says that\n                               run has landed — finished, and its result branch on the base\n                               branch — and start from the base as it is then; once per run.\n                               --from \u003CrunId>: hand this run what that run concluded, as context,\n                               and wait on nothing: landed or still in flight, it is read once,\n                               when this run's steps are filed, and said not landed, not judged\n                               until it has landed; once per run.\n                               --gate \u003Cstep>: that step's closeout waits for nemia approve before\n                               anything after it is filed, as a step's own gate: true does; once\n                               per step. An inbox card's gate: \u003Cstep> is one more\n                               --model \u003Cstep>=\u003Cmodel>: that step runs on the model in this run alone;\n                               its other steps, and every other run, keep their agents' models. Once\n                               per step, over an inbox card's model: for the same step. Refused, naming\n                               the models that keep it, where the step's harness refuses the model or\n                               its permission mode is dropped on it\n                               a file that opens with a frontmatter block of workflow: and after: starts\n                               its body, on its workflow: unless --workflows is given; its after: needs\n                               an --after for each. Any other key is refused by its name, never taken\n                               as the brief: file an inbox card with nemia inbox add \u003Cfile>\n                               --input \u003Cfile>: copied into the run's own inputs\u002F directory, and named\n                               on every step's brief under what was asked for; once per file\n                               --card \u003Cname>: an inbox card, as nemia inbox lists it, copied as it\n                               reads now into the run's inputs\u002Fcards\u002F\u003Cname>.md and named on every\n                               step's brief beside the files; once per card. A name with no card is\n                               refused before anything is written\n                               --factory \u003Cname>: the workflow must be attached to the factory, whose\n                               hires fill its steps, and a step left with no agent is refused (nemia\n                               factory --help). Without it, the run is started in the factory whose\n                               repository holds --repo, else the daemon's own repository, when one\n                               factory alone does; none, or two, is refused, naming the fix. In a\n                               factory the run works on the repository it names for the workflow\n                               (nemia factory attach \u003Cfactory> \u003Cworkflow> --repo \u003Cname>), else its\n                               only one: --repo only finds the factory, and the start says so where\n                               it differs\n                               The default factory is history only: --factory default is refused\n                               a card of the inbox (nemia inbox --help), by its name, starts its body,\n                               or --brief \"\u003Cbrief>\" in its place, and stays in the inbox, listed as\n                               started: --workflows, a name, is its workflow: unless given, --factory\n                               its factory:, and a card that names no factory either way is refused;\n                               --repo is refused: its factory places the run\n                               on the port: start, inbox_start\n  nemia inbox add \"\u003Cidea>\"|\u003Cfile> [--workflow \u003Cname>] [--factory \u003Cname>] [--label \u003Ckey=value>]...\n                               park an idea as a card in the inbox, named from its first line\n                               on the port: inbox_add\n  nemia inbox queue \u003Cname> [--workflow \u003Cw>] [--off]\n                               mark a card to start on its own: the tick starts the oldest queued\n                               card once a run slot is free; --off takes the mark back\n                               on the port: inbox_queue, inbox_unqueue\n  nemia inbox from \u003Cname> \u003Crun>... | --off\n                               set the runs a card is handed as context (from:), read and\n                               never waited on; --off takes them off\n                               on the port: inbox_from\n  nemia inbox edit \u003Cname> \"\u003Ccard>\"|\u003Cfile>\n                               replace a card's body and keys with a card's text, typed or a file;\n                               the card keeps its name, and its labels: nemia label inbox edits those\n                               on the port: inbox_edit\n  nemia label run \u003CrunId> \u003Ckey=value|key->... [--overwrite]\n                               edit a run's labels: key=value sets a key, key- removes it; changing a\n                               value takes --overwrite. nemia ls -l selects by them, and the feed's\n                               lines written after carry them\n                               on the port: label_run\n  nemia label inbox \u003Cname> \u003Ckey=value|key->... [--overwrite]\n                               edit an inbox card's labels: key=value sets a key, key- removes it;\n                               changing a value takes --overwrite. nemia inbox -l selects by them, and\n                               the run the card starts is created with them\n                               on the port: label_inbox\n  nemia route \u003CrunId> \u003Ccondition> \u003Croute>\n                               route a condition by hand: the next tick takes your route over the\n                               condition's own answer, for the try it is on, and files the nodes\n                               whose when: names it. Refused once a node after it is filed or sent:\n                               nemia cancel \u003CrunId>, and start again.\n                               on the port: route\n  nemia refresh \u003CrunId> [\u003Cstep>] [--base]\n                               take the new revisions: a run keeps the workflow, agents, skills and\n                               factory rules it resolved when it started, and this pins it to the\n                               ones they hold now, so every filing after it — a new step, a retry,\n                               a round — is filed on them. What has already been filed is not re-run,\n                               and the steps the run has are neither added to nor removed. Refused of\n                               a run that has filed nothing, which resolves afresh at its first filing\n                               anyway, of one already on the newest revision of every kind, and of one\n                               whose factory follows revisions (nemia factory revisions), whose every\n                               filing takes them with no refresh;\n                               --base: take the base instead — a loop's work that no longer merges\n                               cleanly goes back to merge it, once, on the next tick, on the round it\n                               is on, which the merge does not spend; the step is the one that closes\n                               the loop, left out where the run has one loop\n                               on the port: refresh\n  nemia say \u003CrunId> \u003Cstep> \"\u003Cwords>\"\n                               tell a step's running session something, through its harness: the\n                               engine journals the words on the step's task, nemia ls notes the step\n                               said to, and nemia decide prints them under it. A try that continues\n                               the session is handed them again, as ## Said to your session.\n                               Words to a step not filed yet (waiting on --after, a hold or an\n                               earlier step) are kept, not delivered: its first try is handed them,\n                               as ## Said before you started, and no try after it.\n                               Words are from whoever holds the token sent: a session its run filed\n                               speaks as its agent (WF_SESSION_TOKEN), an agent given a token of its\n                               own (nemia token add \u003Cname> --agent \u003Cagent>) as that agent. An\n                               agent's words reach the step under its name, never as yours\n                               on the port: say\n  nemia agents run yours \u003Cname> --brief \u003Cwords>\n                               an agent of yours tried on a short brief, in a folder of its own,\n                               capped at $1\n                               on the port: agents_run\n  nemia hold \u003CrunId> [--until \u003CrunId> ...] [--why \"\u003Cwords>\"] [--by \u003Cname>]\n                               file nothing more of a run until other runs land, as --after would\n                               have: what is in flight finishes, nothing new is filed, and nothing\n                               is asked of you — nemia ls and nemia decide say held until \u003Crun> lands, and\n                               what a step asks is asked once the hold lifts. The tick lifts it when\n                               every run named has landed. With no --until, the run is held until a\n                               person lifts it (--off), and no tick lifts it: nemia ls and nemia\n                               decide say held until a person lifts it. Refused of a run cancelled,\n                               pruned, finished or already held, and of a run named that is unknown,\n                               the run itself, or one that already waits on it, through --after or a\n                               hold.\n                               --why: the reason, shown beside held until \u003Crun> lands wherever it is\n                               said; --by: who puts the hold (default: your OS user).\n                               nemia hold \u003CrunId> --off: lift the hold now, and the next tick goes on as\n                               if it had never been put\n                               on the port: hold, unhold\n  nemia pause [--why \"\u003Cwords>\"] [--by \u003Cname>] [--timeout \u003Cd> | --now [--yes]]\n                               pause the store: no pass starts new work — no step, try, round or\n                               check filed, no queued task admitted, no queued card started — and\n                               every session, job or attached, parks at its tool call's end: a call\n                               still in flight once the timeout has passed is cut. A check runs on, its\n                               end recorded; a landing whose check has ended commits or aborts.\n                               Pausing: N sessions finishing a tool call, cut in M min while they\n                               do, Paused once none does; nemia ls says each step paused.\n                               nemia start still writes a run, and approvals, retries and answers are\n                               kept, not acted on: the pass after nemia resume files what waits. nemia ls,\n                               nemia watch --once, nemia start and nemia up say it, and nemia ls names what finishes.\n                               Refused while a pause stands.\n                               --why: the reason, said wherever the pause is; --by: who pauses it\n                               (default: your OS user)\n                               --timeout: how long a call in flight is given to end, for this pause\n                               (90s, 10m, 2h; from 1s to 23h); default: the store's, nemia pause-timeout\n                               --now: pause at once, the board's Stop now — each step session at work,\n                               but a landing's, has its call cut at once and parks, going on as the\n                               same try on nemia resume;\n                               each check is asked to stop and run again. A pause that stands is kept.\n                               --yes: with --now, act; without it, --now only says what it would\n                               cut\n                               on the port: pause\n  nemia resume\n                               resume the store: each session parked goes on as the same try, and the\n                               next pass files what waits\n                               on the port: resume\n  nemia subscribe \u003Cname> \u003Ctask> [-l \u003Ckey>=\u003Cvalue>]... [--digest-every \u003Cminutes>] [--at-least \u003Clevel>]\n                               give the engine's long-lived task \u003Ctask> a subscription to the feed,\n                               named \u003Cname>, from its end: a line one of its selectors takes is said to\n                               it at once, and the rest folded into a digest said with the next wake,\n                               or after --digest-every minutes (default 60) with none.\n                               -l: the lines it takes, given again for another (default:\n                               waits=person, the lines that wait on a person). Refused for a task not\n                               long-lived.\n                               --at-least: only the lines of runs whose factory's level is at least\n                               this (manual, semi-auto, auto), read as each line is delivered.\n                               on the port: subscribe\n  nemia unsubscribe \u003Cname>\n                               take the subscription \u003Cname> away: its task is told nothing more of the feed\n                               on the port: unsubscribe\n  nemia policy propose \u003Cagent> --file \u003Cpath>|-\n                               the agent's proposal: rules from your custom instruction, as JSON,\n                               with the words the machinery cannot hold kept as guidance; none is in\n                               force until you confirm it (nemia policy confirm)\n                               on the port: policy_propose\n  nemia daemon [--json]\n                               the store's daemon, started where none ticks it (as any command that\n                               ticks starts it, unless NEMIA_DAEMON_START=off): its state, its pid and\n                               the url the api serves at, from its lease. --json: the whole answer\n                               on the port: daemon\n```\n\n## Prompts the person\n\nEach asks the person at a prompt every time it runs: say what it does before you run it.\n\n```text\n  nemia inbox rm \u003Cname>\n                               drop a card without starting it\n                               on the port: inbox_rm\n  nemia factory add \u003Cname> --repo \u003Cpath|url> [--repo \u003Cpath|url> ...] [--no-trust]\n                               file a factory: a lowercase slug naming one or more directories or git URLs,\n                               each URL cloned, each trusted for its steps unless --no-trust\n                               on the port: factory_add\n  nemia factory apply \u003Cname> --template \u003Cname|folder|archive> --repo \u003Cpath|url> [--repo \u003Cpath|url> ...] [--set \u003Ckey>=\u003Cvalue> ...] [--no-trust]\n                               make a factory from a template that passes its check, its parameters\n                               given with --set; the daemon renders its tree\n                               on the port: factory_apply\n  nemia factory adopt \u003Cname> --template \u003Cname|folder|archive> [--set \u003Ckey>=\u003Cvalue> ...] --plan \u003Cdigest>\n                               take a hand-made factory over by a template, in place, as nemia factory\n                               diff previewed it: its name, cards, runs and stats stay\n                               on the port: factory_adopt\n  nemia factory upgrade \u003Cname> --to \u003Cx.y.z> [--template \u003Cname|folder|archive>] [--set \u003Ckey>=\u003Cvalue> ...] [--plan \u003Cdigest>] [--downgrade]\n                               upgrade a templated factory: its file wants --to, and the daemon merges\n                               the template's change into its tree once each conflict is resolved;\n                               with --to the version applied and --set, a change of its parameters\n                               alone, filed with the --plan nemia factory diff printed for it\n                               on the port: factory_upgrade\n  nemia factory resolve \u003Cname> \u003Cconflict> --ours|--theirs|--file \u003Cpath>\n                               choose for one conflict of a templated factory's upgrade or switch: your\n                               tree, the template (none: drop the file), or the text of a file; once\n                               each conflict of a workflow, an agent, a skill, a hire or the rules and\n                               instruction has one, that object is proposed as they compose it, to\n                               approve (nemia proposals approve \u003Cp> -n \u003Cname>); a carried file stays\n                               a choice; the daemon merges once each has one and each proposal applied\n                               on the port: factory_resolve\n  nemia factory attach \u003Cfactory> \u003Cworkflow> [--repo \u003Cname>]\n                               offer a workflow in a factory, by its folder's name: a factory offers\n                               only the workflows attached to it; --repo names the repository its runs\n                               work on, by folder name, required where the factory has several\n                               on the port: factory_attach\n  nemia factory detach \u003Cfactory> \u003Cworkflow>\n                               take a workflow back, once no hire fills its steps\n                               on the port: factory_detach\n  nemia factory hire \u003Cfactory> \u003Cagent> --step \u003Cworkflow>\u002F\u003Cstep>\n                               hire an agent into a workflow's step: one a step, a second replaces it\n                               on the port: factory_hire\n  nemia factory fire \u003Cfactory> --step \u003Cworkflow>\u002F\u003Cstep>\n                               take a step's hire back\n                               on the port: factory_fire\n  nemia factory schedule \u003Cfactory> \u003Cworkflow> --cron \u003Cexpr> --brief \u003Ctext> [--tz \u003Czone>]\n                               start a run of an attached workflow at each fire of a cron, by the\n                               daemon, at every level; a second replaces the first\n                               on the port: factory_schedule\n  nemia factory unschedule \u003Cfactory> \u003Cworkflow>\n                               take a workflow's schedule back: the daemon starts it no more\n                               on the port: factory_unschedule\n  nemia factory server add \u003Cfactory> \u003Cserver> --url \u003Caddress> [--transport http|sse] [--header \u003Cname>=\u003Cvalue> ...] [--client-id \u003Cid>] [--callback-port \u003Cport>]\n                               record a remote MCP server for a factory: its steps' sessions are\n                               handed it, and an agent `mcp:` line names it\n                               on the port: factory_server_add\n  nemia factory server rm \u003Cfactory> \u003Cserver>\n                               take a server off a factory's record\n                               on the port: factory_server_rm\n  nemia factory volume add \u003Cfactory> \u003Cpath>:ro|:rw\n                               mount a folder of this machine for every local step of a factory's runs,\n                               read-only (:ro) or writable (:rw); a writable one is not a run's workspace\n                               on the port: factory_volume_add\n  nemia factory volume rm \u003Cfactory> \u003Cpath>\n                               take a volume off a factory, by its path\n                               on the port: factory_volume_rm\n  nemia factory setting set \u003Cfactory> \u003Charness>.\u003Ckey> \u003Cjson>\n                               give every session of a factory's runs on a harness one of its settings,\n                               claude.attribution among them; its value is JSON\n                               on the port: factory_setting_set\n  nemia factory setting rm \u003Cfactory> \u003Charness>.\u003Ckey>\n                               take a harness setting off a factory, by its key\n                               on the port: factory_setting_rm\n  nemia factory cost-line set \u003Cfactory> \u003Cworkflow> \u003Cusd>\n                               set what a run of a workflow may cost before its gate says so, in dollars:\n                               a gate whose caps pass it is said, and never approved unasked at auto\n                               on the port: factory_cost-line_set\n  nemia factory cost-line rm \u003Cfactory> \u003Cworkflow>\n                               take a workflow's cost line off a factory\n                               on the port: factory_cost-line_rm\n  nemia factory instruct \u003Cname> --file \u003Cpath>|- | --clear\n                               write the instruction the store keeps for a factory (- for stdin), what\n                               Nemia is to know of it and no step is told, or remove it with --clear\n                               on the port: factory_instruct, factory_instruct_delete\n  nemia factory revisions \u003Cname> follow|pinned\n                               whether a factory's runs follow merged revisions at each filing (follow)\n                               or keep them until nemia refresh (pinned, the default)\n                               on the port: factory_revisions\n  nemia factory trust \u003Cname>\n                               trust a factory's places for the harnesses its steps run on, so each\n                               place's own settings apply to them\n                               on the port: factory_trust\n  nemia factory rm \u003Cname>\n                               remove a factory's file, and the instruction and the rules the store\n                               keeps for it\n                               on the port: factory_rm\n  nemia template export \u003Cfolder> --from \u003Crepo> [--factory \u003Cname>]\n                               write a template from a repository and a factory, and say where each\n                               part came from\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia template pack \u003Cfolder> [--out \u003Cfile>]\n                               pack a template's folder that passes its check into one archive, and print\n                               the digest of its content\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia template unpack \u003Carchive> \u003Cdir>\n                               write the folder a template's archive holds at \u003Cdir>\u002F\u003Cname>, each file's\n                               bytes and mode as packed\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia agents write (yours | factory) [\u003Cname>] --file \u003Cname>.md [--factory \u003Cname>] [--version \u003Cv>]\n                               an agent of yours, or of a templated --factory, written from its file:\n                               \u003Cname> is the agent it changes or renames, at the --version it was\n                               read at; none, a new one\n                               on the port: agents_write, agents_replace\n  nemia agents copy \u003Cfile> [--as \u003Cname>] [--keep \u003Cpart>]...\n                               an agent file copied into your library, with its schema and skills;\n                               a part your library holds otherwise stops it: --as or --keep\n                               on the port: agents_copy\n  nemia agents rm yours \u003Cname> --version \u003Cv>\n                               an agent of yours removed, at the --version it was read at\n                               on the port: agents_rm\n  nemia skills write (yours | factory) [\u003Cname>] --file \u003CSKILL.md> [--factory \u003Cname>] [--version \u003Cv>]\n                               a skill of yours, or of a templated --factory, written from its\n                               SKILL.md: \u003Cname> is the skill it changes, at the --version it was\n                               read at; none, a new one\n                               on the port: skills_write, skills_replace, skills_copy\n  nemia cancel \u003CrunId>\n                               stop what it has in flight; file nothing more\n                               on the port: cancel\n  nemia rm \u003CrunId> [\u003CrunId> ...] [--delete-branches [--force-unmerged]]\n                               forget runs that are over, a line each: a run's worktrees removed\n                               through git, what one held uncommitted committed first, then its\n                               record and every file under WF_ROOT keyed by its id. Refused while\n                               a task of it is in flight, anything is due to be filed or asked\n                               for, or a run waiting on it has not started. Its branches are kept:\n                               --delete-branches deletes those merged into the run's base branch,\n                               and --force-unmerged beside it the others, with their work. Then\n                               the engine forgets the run's tasks: their journals and session\n                               files; one not settled yet is forgotten by the next passes\n                               on the port: rm\n  nemia prune [\u003CrunId> ...] [--unmerged] [--artifacts] [--yes]\n                               free the space runs that are over hold: their worktrees, removed\n                               through git, and nothing else — each run's record, branches and\n                               files are kept, and it stays on nemia ls, nemia decide and the board,\n                               marked pruned. --artifacts removes its files too (runs\u002F\u003Cid>\u002F\n                               artifacts\u002F: screenshots, logs), of a run whose worktrees are\n                               already gone as well; its record and branches are still kept.\n                               Lists what it would remove, with its size, and removes it on\n                               --yes, or on y at the prompt. With no run, every run. Refused\n                               while a run is not over, owes a hand-off, is a folder run, or has\n                               a worktree held, on no branch, or with changes no commit has;\n                               and a finished run not merged unless --unmerged, with --artifacts\n                               as without: its files are what a person reads to decide it\n                               on the port: prune\n  nemia pause-timeout [\u003Cd>]\n                               the store's pause timeout: how long nemia pause, nemia service restart and\n                               nemia upgrade give each tool call in flight to end before it is cut; bare,\n                               the timeout in force (10m by default); given one, sets it: digits and s, m\n                               or h, from 1s to 23h, as 90s, 10m or 2h. nemia pause --timeout \u003Cd> gives one\n                               pause a timeout of its own\n                               on the port: pause-timeout, pause_timeout_set\n  nemia policy confirm \u003Cagent> \u003Cid> --words '\u003Cwords>'\n                               put in force a rule the agent proposed, or the default again, as its\n                               words read: from then on, at auto, the answer it names is given unasked\n                               where it holds. Refused once the rule reads otherwise. Your first confirm\n                               keeps the default beside it. Asked of you every time\n                               on the port: policy_confirm\n  nemia policy rm \u003Cagent> \u003Cid>\n                               take a rule out of force, the default among them: what it allowed is\n                               asked of you again. Asked of you every time\n                               on the port: policy_rm\n  nemia service install [--env KEY=VALUE ...] [--size auto] [--bind \u003Caddr> [--public]] [--name \u003Cname>] [--take-over \u003Cpid>]\n                               install the daemon as a login service: a systemd user unit on Linux,\n                               a launchd agent on macOS. It runs the api of this checkout with this\n                               node, WF_ROOT and PATH, and each --env, written once; it restarts on\n                               failure and starts at login. Started now unless another daemon holds\n                               the store; --take-over \u003Cpid> pauses the store, waits for the calls\n                               in flight of its sessions, stops that pid and starts the service\n                               (checks and jobs run on, as nemia service restart --help says).\n                               --size auto: the sessions at once and each one's memory limit and\n                               request, read off this machine (WF_MAX_CONCURRENT, WF_MEMORY_LIMIT,\n                               WF_MEMORY_REQUEST; neither memory one on macOS), each but one --env\n                               gives; the floor kept, WF_MEMORY_RESERVED, is its default (an eighth\n                               of the machine, 1 GiB at least). On systemd it also writes\n                               \u003Cname>-sessions.slice, its MemoryMax at the machine less that floor,\n                               and points WF_CGROUP_PARENT at it, so the sessions together are\n                               held to what the machine allocates (none written where --env gives\n                               WF_CGROUP_PARENT: a line names its set-property). --bind \u003Caddr>: the\n                               address the daemon listens on (NEMIA_BIND), and the store marked\n                               exposed (WF_EXPOSED=1), so every caller presents a token: nemia\n                               pairings approve and nemia board link hand out the first. A public or\n                               wildcard address is refused unless --public acknowledges it\n                               (NEMIA_BIND_PUBLIC=1); nemia service addresses lists this machine's.\n                               --name: the service name (nemia; required for a store that is not\n                               the default)\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service uninstall [--name \u003Cname>]\n                               stop the login service, disable it and remove its file and the\n                               sessions' slice --size auto wrote; a daemon started by hand is left\n                               alone\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service restart [--name \u003Cname>] [--timeout \u003Cd>]\n                               restart the service's daemon on the code it runs: pause the store, give\n                               the daemon's sessions up to the timeout to end their tool calls\n                               (--timeout \u003Cd>, else nemia pause-timeout, 10m by default), stop it (up to\n                               90s, and until the unit has ended its process), start the service (up to\n                               60s), and resume once the new daemon has completed its first pass;\n                               refused, naming the log, when that pass fails, and naming the command\n                               that starts the service when it is not running.\n                               Jobs and checks, a landing waiting on CI among them, run on and are not\n                               waited for; each session parks and resumes as the same try. Each wait is\n                               said on stderr as it waits. A pause a person put stands; the restart's\n                               own ends with the command: killed, the next pass of the daemon lifts it.\n                               Starts the service where nothing ticks; a daemon started by hand is\n                               named with --take-over's line\n                               CLI only: it restarts the person's login service, on the machine the command runs on\n  nemia upgrade [\u003Ctag>] [--repo \u003Cpath>] [--name \u003Cname>] [--timeout \u003Cd>]\n                               on an archive install, download the tag's archive (the newest where\n                               none is named), check it against its SHA256SUMS and run its nemia\n                               version, into the store's releases\u002F\u003Ctag>; on a release built from a\n                               repository, build the tag from a clean export into releases\u002F\u003Ctag>:\n                               npm ci, npm run build and npm run release:check there, on a trial\n                               store; then point releases\u002Fcurrent at it and restart the daemon onto\n                               it where the login service runs the pointer: jobs run on, and the\n                               daemon's sessions are given up to the timeout to end their tool calls\n                               (--timeout \u003Cd>, else nemia pause-timeout), park, and resume as the same try on\n                               the new release; then say each piece the new release reads missing\n                               that a run waits on, and each of the machine and its service, with\n                               its fix, or that none of those is. A tag built already is switched\n                               to without a rebuild, so an older tag rolls back. --repo:\n                               the repository the tag is in (the running release's source);\n                               --name: the service (nemia). Then it removes each release that is\n                               not current, not among the two newest, and that no session or\n                               running nemia runs on, the folder of each build whose upgrade died,\n                               and each build log older than 30 days, as the daemon does every five\n                               minutes. A tag older than the oldest release that can read the store\n                               (store.json) is refused before anything is fetched or built. It\n                               upgrades this machine's install, whatever the context: the server a\n                               context names is upgraded from here with nemia server upgrade [\u003Ctag>],\n                               through its api (not nemia service, this machine's login unit)\n                               on the port: upgrade\n  nemia frontmatter rewrite [\u003Cfolder>...] [--check]\n                               rewrite as YAML each markdown file's frontmatter still in the former\n                               line form, under the folders given or the store's own. --check: write\n                               nothing, and exit 1 where a rewrite is due\n                               CLI only: it rewrites files on the machine nemia runs on, folders a person names among them\n  nemia workflows write (yours | factory) \u003Cname> --folder \u003Cdraft> [--factory \u003Cname>] [--version \u003Cv>]\n                               a workflow folder of yours, or of a templated --factory, written from\n                               a draft folder, its steps, README, scripts and schemas, leaving a file\n                               it does not carry; one there already, at the --version it was read at\n                               on the port: workflows_write, workflows_replace\n  nemia workflows rm yours \u003Cname> --version \u003Cv>\n                               a workflow folder of yours removed, at the --version it was read at\n                               on the port: workflows_rm\n```\n\n## Answers what waits\n\nEach answers what waits on the person. Bring the matter first, with the `bring-what-waits` skill, then run the verb they choose, at a shell with their words as `--message`, or its tool on the port with them as its message: its prompt is their decision, asked once. `retry` asks only of a step that waits on the person, or with `--budget`.\n\n```text\n  nemia retry \u003CrunId> [\u003Cstep>] [--continue | --fresh | --model \u003Cm> | --recheck | --round] [--answer \u003Cn>=\u003Cletter>]... [--message \"\u003Canswer>\"] [--budget \u003Camount>]\n                               try a step waiting on you again, once, on the next tick:\n                               --continue (default) continues its session, told what happened;\n                               --fresh a new session on its branch; --model \u003Cm> the same, on m;\n                               --recheck runs its check again, and files no session;\n                               --round: once more — a loop out of rounds goes round again, once,\n                               the step it loops to continued, then a new reviewer; the step is the\n                               one that closes the loop, left out where the run has one loop;\n                               --message \"\u003Canswer>\" goes into the try's brief, as ## Your answer:\n                               for a step that signed off blocked, the answer to its needs — or\n                               what you have to say to any step you try again; not with --recheck\n                               or --round. The next argument is the answer, whatever it starts with;\n                               an answer that is only one of these flags' names goes as --message=\u003Canswer>;\n                               --budget \u003Camount>: for a step with a budget that waits on you, its\n                               cap raised to that many dollars, above its cap and what it has cost;\n                               with --round, the cap of a step of the loop over its budget;\n                               --answer \u003Cn>=\u003Cletter>: answer question n, numbered as nemia decide\n                               \u003CrunId> prints them, with the option its letter names; once for each\n                               question: \"\u003Cn>=a,c\" chooses several where it takes several,\n                               \"\u003Cn>=a: \u003Cwords>\" adds your own words, \"\u003Cn>: \u003Cwords>\" says words\n                               alone; --message is then the note under your answers\n                               on the port: retry\n  nemia approve \u003CrunId> [\u003Cstep>] [--answer \u003Cn>=\u003Cletter>]... [--budget \u003Cstep>=\u003Camount>]... [--message \"\u003Cwords>\"] | \u003CrunId> \u003Cstep> --off\n                               let a step waiting at its gate hand on: what waits on it is filed on\n                               the next tick, and every step after the gate is handed your words\n                               under ## Approved. To send it back instead, with what to change:\n                               nemia retry \u003CrunId> \u003Cstep> --message \"\u003Cobjection>\"\n                               Of a loop out of rounds — the step that closes it, or the run alone\n                               where it has one loop — enough: the loop ends on your words, with its\n                               findings open, --message \"\u003Cwhy>\" required. What waits on the looping\n                               step is filed as if it had said again: false, and every step after\n                               the loop is handed your words and those findings, under the heading\n                               of a loop ended as it stands.\n                               To go round once more instead: nemia retry \u003CrunId> [\u003Cstep>] --round\n                               --answer \u003Cn>=\u003Cletter>: answer question n, numbered as nemia decide\n                               \u003CrunId> prints them, with the option its letter names; once for each\n                               question: \"\u003Cn>=a,c\" chooses several where it takes several,\n                               \"\u003Cn>=a: \u003Cwords>\" adds your own words, \"\u003Cn>: \u003Cwords>\" says words\n                               alone; --message is then the note under your answers\n                               --budget \u003Cstep>=\u003Camount>: the cap of a step after the gate, in\n                               dollars, over what the step proposed for it; once for each step.\n                               Your approval makes the budgets the step proposed the caps of the\n                               steps after the gate, and auto never approves unasked a cap above\n                               a step's own budget\n                               --off: take back the approval of a step that stands approved while\n                               nothing it approved has been filed — an ending waiting on the base\n                               branch, held, or not yet ticked. The step waits at its gate again\n                               on the port: approve\n```\n","# The verbs\n\nEvery verb `nemia` serves you, by who answers for it, with its usage and help as `nemia --help` prints them. `nemia \u003Cverb> --help`\nsays more of one that has a manual. With a shell, run `nemia \u003Cverb>`; on the MCP port, call the tool its `on the port` line names,\nwhose schema says its arguments. A verb said `CLI only` runs at a shell alone, for the reason given.\n\nGenerated from the command table, never edited by hand: from the repository's root, `node -e \"import('.\u002Fpackages\u002Fagent-orchestrator\u002Fsrc\u002Fnemia\u002Fverbs.verb.ts').then((m) => process.stdout.write(m.verbsSkill()))\" > packages\u002Fagent-orchestrator\u002Fnemia\u002Fskills\u002Fverbs\u002FSKILL.md`.\n\n## Reads\n\nThey write nothing: run them unasked, as often as you need. One that sets with some words prompts the person given them.\n\n```text\n  nemia concluded \u003CrunId>\n                               what a run started with --from \u003CrunId> would be handed of it, now\n                               on the port: concluded\n  nemia inbox\n                               the cards still in play: name, first line, workflow when set,\n                               and started as \u003CrunId>, \u003Cphase> (nemia inbox --help); --all: those whose\n                               run ended too; -l: the cards whose labels match; --show-labels, -L \u003Ckey>:\n                               their labels on each line; -o wide: every column\n                               on the port: inbox, inbox_card\n  nemia factory\n                               the store's factories: the default first — every run and card that\n                               names no factory, read, never written — then each by name with its\n                               repositories (nemia factory --help)\n                               on the port: factory\n  nemia factory ls\n                               the same, by its name\n                               on the port: factory_ls\n  nemia factory diff \u003Cname> --to \u003Cx.y.z> [--template \u003Cname|folder|archive>] [--set \u003Ckey>=\u003Cvalue> ...] | \u003Cname> --template \u003Cname|folder|archive> [--set \u003Ckey>=\u003Cvalue> ...]\n                               preview a templated factory's upgrade: what merges cleanly, and each\n                               conflict with its sides; onto a template of another name, a switch\n                               with no common base; or a hand-made factory's takeover by a\n                               template, and the plan nemia factory adopt takes; nothing is written.\n                               To change a parameter alone: --to the version applied, and\n                               --set \u003Ckey>=\u003Cvalue> for each; it prints the line that files it\n                               on the port: factory_diff\n  nemia factory servers \u003Cfactory>\n                               a factory's own MCP servers, each with the state its harness reports\n                               on the port: factory_servers\n  nemia factory rules \u003Cname> [--file \u003Cpath>|- | --clear]\n                               a factory's rules and policies: words told to every agent hired into\n                               it, after its custom instructions and before the step's own note,\n                               kept by the store with the factory. Bare, prints them; --file writes\n                               them (- for stdin); --clear removes them. Words only: a frontmatter\n                               block is refused. A run keeps the rules it was started with\n                               reads bare; prompts the person given --file, --clear\n                               on the port: factory_rules, factory_rules_put, factory_rules_delete\n  nemia factory instruction \u003Cname> [--path \u003Cpath> | --store | --clear]\n                               where a factory's instruction is kept, what Nemia is to know of the\n                               factory (its agents are told its rules: nemia factory rules): bare,\n                               prints it; --path reads it from a file in the factory's first folder,\n                               --store keeps it in the store, --clear reads NEMIA.md in the first\n                               folder again\n                               reads bare; prompts the person given --path, --store, --clear\n                               on the port: factory_instruction, factory_instruction_put, factory_instruction_delete\n  nemia template lint \u003Cfolder|archive>\n                               check a template's folder or archive: every problem at once, each at its\n                               file and line\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia ls [\u003CrunId>] [-l \u003Ckey>=\u003Cvalue>] [--show-labels] [-L \u003Ckey>] [-o wide]\n                               runs and their steps — a run waiting on another says waiting on \u003Crun>,\n                               one the cap on runs in flight holds waiting for a run slot (2 of 2 in\n                               flight); what each step's tries and the run have cost, where measured;\n                               last, whether a nemia up ticks the store and when its last pass was;\n                               first, while the store is paused, Pausing: N finishing and each\n                               session, check or landing at work, or Paused once none is, then each\n                               check it holds, a landing with its merge staged in the repository's\n                               own working tree (MERGE_HEAD) among them; a step whose check it holds\n                               says waiting on the pause. -l: the runs whose labels match (team=core,\n                               given again for another); --show-labels: each run with its labels;\n                               -L \u003Ckey>: that label beside each run, given again for another; -o\n                               wide: every column, each step among them\n                               on the port: ls, ls_run\n  nemia show \u003CrunId> [--raw]\n                               the run as one object (nemia\u002Fv1 Run); --raw: its record, and\n                               what each step was told\n                               on the port: show\n  nemia closeout \u003CrunId> \u003Cstep> [--try \u003Cn>]\n                               what a step's try signed off with: its latest, or --try \u003Cn>\n                               on the port: closeout\n  nemia logs \u003CrunId> \u003Cstep>\n                               that step's session output\n                               on the port: logs\n  nemia stats [--from \u003Cday>] [--to \u003Cday>] [--by agent|step|factory|workflow|repo] [--agent|--step|--factory|--workflow|--repo \u003Cv>] [-l \u003Ckey>=\u003Cvalue>] [--line|--json]\n                               what every run in the store adds up to, by UTC day: rounds to\n                               acceptance of each loop: step, landings of the workspace: base step\n                               and on which try, cost and time per landed run, and each step's\n                               tries, time, cost, checks and gate waits; first the last 14 days\n                               against the 14 before. A number nothing measured reads —, never 0.\n                               --line: one line, for a recap; --json: the whole answer; -l: the\n                               runs whose labels it takes, as nemia ls -l takes them\n                               on the port: stats\n  nemia passes [--last \u003Cn>] [--json]\n                               what the recent passes of the tick cost: their median and slowest\n                               time, a histogram of their times, where the time went phase by\n                               phase, and what the last saw — the runs in the store, its journal\n                               reads, the sessions queued and admitted. --last \u003Cn>: the last n\n                               passes only; --json: the whole answer\n                               on the port: passes\n  nemia series [--range 24h|7d|30d|all] [--json]\n                               the system's series over a range: a pass's time and its phases, its\n                               journal reads, the runs in the store, the sessions queued and admitted\n                               and the longest wait, the load, and the quota used. --range: 24h (a\n                               point a minute, the default), 7d or 30d (an hour), all (a day);\n                               --json: the whole answer\n                               on the port: series\n  nemia agents [--repo .]\n                               the agents a step may name, the repository's then yours: what each is\n                               for, what it says it can do (an entry for each skill, read off its\n                               SKILL.md), and the skills it is given (its skills:)\n                               on the port: agents\n  nemia agents card \u003Cname> [--repo .] [--from repo|yours]\n                               that agent in the shape of an A2A Agent Card, JSON on stdout and\n                               nothing else — every field the protocol marks REQUIRED but\n                               supportedInterfaces, which is left out because nothing here serves an\n                               agent over A2A. The card describes; it is not an address\n                               on the port: agents_card\n  nemia agents sources [--path \u003Cfolder>]\n                               the agent files the store knows of, to copy into your library:\n                               each repository's, yours, and those of a --path \u003Cfolder>\n                               on the port: agents_sources\n  nemia agents custom \u003Cname> [--file \u003Cpath>|-] [--clear] [--repo .]\n                               your custom instructions for that agent: words you add to it without\n                               rewriting it, kept in \u003Cstore>\u002Fcustom-instructions\u002F\u003Cname>.md and told\n                               after its instructions. Bare, prints them; --file writes them (- for\n                               stdin); --clear removes them. Words only: a frontmatter block is\n                               refused. A run keeps the words it was started with; nemia, the\n                               store's own agent, is told them in her open session\n                               reads bare; prompts the person given --file, --clear\n                               on the port: agents_custom, agents_custom_put, agents_custom_delete, nemia_instructions, nemia_instructions_put, nemia_instructions_delete\n  nemia wait [-l \u003Ckey>=\u003Cvalue>]... [--from \u003Ccursor>] [--for \u003Cduration>]\n                               wait for the feed's next line -l takes, and print it with the\n                               cursor after it: from --from (a cursor a wait or nemia watch printed),\n                               else from the feed's end now. It ends on the first lines -l takes, or\n                               after --for (4m by default, at most 270s) with none, and prints the\n                               cursor either way: wait again from it and no line is lost.\n                               -l: as nemia watch -l takes it, given again for another; none takes\n                               every line.\n                               on the port: wait\n  nemia level [manual|semi-auto|auto] [--factory \u003Cname> [--for \u003Cduration> | --end]]\n                               Nemia, the store's own agent: bare, where she is, each factory's level and\n                               how to reach her; a level sets the store's, which every factory follows\n                               unless it has its own; --factory sets that factory's, --for raises it for\n                               a stretch (90m, 2h, 1d; up to 7d) and --end ends the raise. semi-auto keeps\n                               her session: what waits on you (a gate, a step's question, a loop out of\n                               rounds, a cap) is asked of you, and a write that cannot be undone or\n                               changes your settings or a factory; the rest is hers. auto is semi-auto,\n                               but an answer your confirmed list allows in the run's state is given\n                               unasked (nemia policy \u003Cagent>), in a session of hers filed since auto came:\n                               one kept from before asks every answer until filed again (manual, then\n                               auto once it is off). A factory at manual is left alone, and her session\n                               stops once none is above manual\n                               reads bare; prompts the person given manual, semi-auto, auto, --factory, --for, --end\n                               on the port: level, level_set\n  nemia memory [\u003Cfact>]\n                               what Nemia keeps of the store: bare, her facts' names; given one, its\n                               words. Read only: her session alone writes them\n                               on the port: memory, memory_fact\n  nemia policy \u003Cagent>\n                               the agent's rules at auto: the answers it may give unasked, in force once\n                               you confirm them, and what it proposed from your custom instruction,\n                               with the words the machinery cannot hold, kept as guidance\n                               on the port: policy\n  nemia policy reach \u003Cagent> (--factory \u003Cname> | --run \u003CrunId>)\n                               which steps at their gate the agent's rules in force could ever approve\n                               unasked, in a factory or a run: why each other rule cannot (a step approved\n                               before it runs has done nothing to recommend or check; one with no check\n                               never passed one), what keeps a rule from a run now, and the rules that would\n                               on the port: policy_reach\n  nemia service status [--name \u003Cname>]\n                               the login service and the lease: whether it is installed and runs,\n                               the checkout it runs, who holds the store, and lingering\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service addresses [--json]\n                               this machine's addresses, loopback first, one a line: the address, its\n                               interface and its kind (loopback, private, shared, link-local,\n                               unique-local, public), so --bind takes a private one; --json prints\n                               {\"addresses\":[{\"address\",\"interface\",\"family\",\"kind\"}]}\n                               CLI only: it reads the network interfaces of the machine the command runs on\n  nemia doctor [--json]\n                               every piece this server needs before a run: each harness installed and\n                               signed in, the service and its linger, the bind, the machine, the git\n                               identity, each factory's workflows' and repository's needs, a factory, a\n                               newer release — each in place or missing, with the command or fix that\n                               puts it there. --json: the whole read\n                               on the port: doctor\n  nemia readiness\n                               retired: nemia doctor reads what it read\n                               CLI only: retired: nemia doctor is the read, served on the port\n  nemia token ls\n                               the tokens you minted, by name: who holds each, the factory and the\n                               narrowing of an agent’s, and when it was minted. Never a token itself\n                               on the port: token_ls\n  nemia pairings ls\n                               the pairings the store keeps, newest first: the client that asked, and\n                               whether it waits for your approval, is approved, was claimed or lapsed.\n                               Never a code: approve the one the client shows with nemia pairings approve\n                               on the port: pairings_ls\n  nemia claude-token status\n                               whether your subscription token is kept, its file and its mode, when it\n                               was set and the days left of its year, warned of a month before it ends:\n                               never the token\n                               CLI only: the token is the person's own, kept on the machine the store is on and read there from the terminal; no caller of the port reads or writes it\n  nemia version [--json]\n                               this nemia's release, an install's tag or a checkout, and its folder;\n                               the release the store's releases\u002Fcurrent names; and the release the\n                               daemon runs, from its lease, and whether it promises to keep a running\n                               session on its files; and the sessions at work by release and harness\n                               version, off their records; and a newer release, where the daemon's\n                               release check read one. --json: the whole answer\n                               CLI only: it says which code the command runs from; the api answers its own at GET \u002Fapi\u002Fversion\n  nemia admit [--at-least \u003Clevel>] [--auto \u003Cagent>]\n                               a call's admission: the call a session is about to make, on stdin, answered\n                               ask where it answers what waits on you;\n                               --at-least: deny a call acting on a run or a factory below the level;\n                               --auto: let through, for a run of a factory at auto, an answer the agent's\n                               rules you confirmed allow in the run as it stands (nemia policy \u003Cagent>)\n                               CLI only: a session's hook asks it before each call, on its own machine\n  nemia decide \u003CrunId> [--sensitive \u003Cglobs>]\n                               what the run produced, and the choice it leaves you (nemia decide --help)\n                               on the port: decide\n```\n\n## Yours\n\nThey write, and are yours to run unasked, on your judgement: say what you did.\n\n```text\n  nemia start \u003Cbrief.md|card|\"text\"> --workflows \u003Cname> [--repo .] [--no-worktree] [--after \u003CrunId> ...] [--from \u003CrunId> ...] [--input \u003Cfile> ...] [--card \u003Cname> ...] [--gate \u003Cstep> ...] [--model \u003Cstep>=\u003Cmodel> ...] [--factory \u003Cname>] [--brief \"\u003Cbrief>\"]\n                               start a run of a workflow from a brief, a text, or a card of the inbox\n                               --workflows \u003Cname>: the workflow, by its name, as its factory offers\n                               it (nemia workflows lists them); a card's own workflow: where it names\n                               one. A path to a workflow folder, and a folder of cards, are refused:\n                               file each card (nemia inbox add \u003Ccard>) and start it by its name\n                               --after \u003CrunId>: file nothing of this run until nemia decide says that\n                               run has landed — finished, and its result branch on the base\n                               branch — and start from the base as it is then; once per run.\n                               --from \u003CrunId>: hand this run what that run concluded, as context,\n                               and wait on nothing: landed or still in flight, it is read once,\n                               when this run's steps are filed, and said not landed, not judged\n                               until it has landed; once per run.\n                               --gate \u003Cstep>: that step's closeout waits for nemia approve before\n                               anything after it is filed, as a step's own gate: true does; once\n                               per step. An inbox card's gate: \u003Cstep> is one more\n                               --model \u003Cstep>=\u003Cmodel>: that step runs on the model in this run alone;\n                               its other steps, and every other run, keep their agents' models. Once\n                               per step, over an inbox card's model: for the same step. Refused, naming\n                               the models that keep it, where the step's harness refuses the model or\n                               its permission mode is dropped on it\n                               a file that opens with a frontmatter block of workflow: and after: starts\n                               its body, on its workflow: unless --workflows is given; its after: needs\n                               an --after for each. Any other key is refused by its name, never taken\n                               as the brief: file an inbox card with nemia inbox add \u003Cfile>\n                               --input \u003Cfile>: copied into the run's own inputs\u002F directory, and named\n                               on every step's brief under what was asked for; once per file\n                               --card \u003Cname>: an inbox card, as nemia inbox lists it, copied as it\n                               reads now into the run's inputs\u002Fcards\u002F\u003Cname>.md and named on every\n                               step's brief beside the files; once per card. A name with no card is\n                               refused before anything is written\n                               --factory \u003Cname>: the workflow must be attached to the factory, whose\n                               hires fill its steps, and a step left with no agent is refused (nemia\n                               factory --help). Without it, the run is started in the factory whose\n                               repository holds --repo, else the daemon's own repository, when one\n                               factory alone does; none, or two, is refused, naming the fix. In a\n                               factory the run works on the repository it names for the workflow\n                               (nemia factory attach \u003Cfactory> \u003Cworkflow> --repo \u003Cname>), else its\n                               only one: --repo only finds the factory, and the start says so where\n                               it differs\n                               The default factory is history only: --factory default is refused\n                               a card of the inbox (nemia inbox --help), by its name, starts its body,\n                               or --brief \"\u003Cbrief>\" in its place, and stays in the inbox, listed as\n                               started: --workflows, a name, is its workflow: unless given, --factory\n                               its factory:, and a card that names no factory either way is refused;\n                               --repo is refused: its factory places the run\n                               on the port: start, inbox_start\n  nemia inbox add \"\u003Cidea>\"|\u003Cfile> [--workflow \u003Cname>] [--factory \u003Cname>] [--label \u003Ckey=value>]...\n                               park an idea as a card in the inbox, named from its first line\n                               on the port: inbox_add\n  nemia inbox queue \u003Cname> [--workflow \u003Cw>] [--off]\n                               mark a card to start on its own: the tick starts the oldest queued\n                               card once a run slot is free; --off takes the mark back\n                               on the port: inbox_queue, inbox_unqueue\n  nemia inbox from \u003Cname> \u003Crun>... | --off\n                               set the runs a card is handed as context (from:), read and\n                               never waited on; --off takes them off\n                               on the port: inbox_from\n  nemia inbox edit \u003Cname> \"\u003Ccard>\"|\u003Cfile>\n                               replace a card's body and keys with a card's text, typed or a file;\n                               the card keeps its name, and its labels: nemia label inbox edits those\n                               on the port: inbox_edit\n  nemia label run \u003CrunId> \u003Ckey=value|key->... [--overwrite]\n                               edit a run's labels: key=value sets a key, key- removes it; changing a\n                               value takes --overwrite. nemia ls -l selects by them, and the feed's\n                               lines written after carry them\n                               on the port: label_run\n  nemia label inbox \u003Cname> \u003Ckey=value|key->... [--overwrite]\n                               edit an inbox card's labels: key=value sets a key, key- removes it;\n                               changing a value takes --overwrite. nemia inbox -l selects by them, and\n                               the run the card starts is created with them\n                               on the port: label_inbox\n  nemia route \u003CrunId> \u003Ccondition> \u003Croute>\n                               route a condition by hand: the next tick takes your route over the\n                               condition's own answer, for the try it is on, and files the nodes\n                               whose when: names it. Refused once a node after it is filed or sent:\n                               nemia cancel \u003CrunId>, and start again.\n                               on the port: route\n  nemia refresh \u003CrunId> [\u003Cstep>] [--base]\n                               take the new revisions: a run keeps the workflow, agents, skills and\n                               factory rules it resolved when it started, and this pins it to the\n                               ones they hold now, so every filing after it — a new step, a retry,\n                               a round — is filed on them. What has already been filed is not re-run,\n                               and the steps the run has are neither added to nor removed. Refused of\n                               a run that has filed nothing, which resolves afresh at its first filing\n                               anyway, of one already on the newest revision of every kind, and of one\n                               whose factory follows revisions (nemia factory revisions), whose every\n                               filing takes them with no refresh;\n                               --base: take the base instead — a loop's work that no longer merges\n                               cleanly goes back to merge it, once, on the next tick, on the round it\n                               is on, which the merge does not spend; the step is the one that closes\n                               the loop, left out where the run has one loop\n                               on the port: refresh\n  nemia say \u003CrunId> \u003Cstep> \"\u003Cwords>\"\n                               tell a step's running session something, through its harness: the\n                               engine journals the words on the step's task, nemia ls notes the step\n                               said to, and nemia decide prints them under it. A try that continues\n                               the session is handed them again, as ## Said to your session.\n                               Words to a step not filed yet (waiting on --after, a hold or an\n                               earlier step) are kept, not delivered: its first try is handed them,\n                               as ## Said before you started, and no try after it.\n                               Words are from whoever holds the token sent: a session its run filed\n                               speaks as its agent (WF_SESSION_TOKEN), an agent given a token of its\n                               own (nemia token add \u003Cname> --agent \u003Cagent>) as that agent. An\n                               agent's words reach the step under its name, never as yours\n                               on the port: say\n  nemia agents run yours \u003Cname> --brief \u003Cwords>\n                               an agent of yours tried on a short brief, in a folder of its own,\n                               capped at $1\n                               on the port: agents_run\n  nemia hold \u003CrunId> [--until \u003CrunId> ...] [--why \"\u003Cwords>\"] [--by \u003Cname>]\n                               file nothing more of a run until other runs land, as --after would\n                               have: what is in flight finishes, nothing new is filed, and nothing\n                               is asked of you — nemia ls and nemia decide say held until \u003Crun> lands, and\n                               what a step asks is asked once the hold lifts. The tick lifts it when\n                               every run named has landed. With no --until, the run is held until a\n                               person lifts it (--off), and no tick lifts it: nemia ls and nemia\n                               decide say held until a person lifts it. Refused of a run cancelled,\n                               pruned, finished or already held, and of a run named that is unknown,\n                               the run itself, or one that already waits on it, through --after or a\n                               hold.\n                               --why: the reason, shown beside held until \u003Crun> lands wherever it is\n                               said; --by: who puts the hold (default: your OS user).\n                               nemia hold \u003CrunId> --off: lift the hold now, and the next tick goes on as\n                               if it had never been put\n                               on the port: hold, unhold\n  nemia pause [--why \"\u003Cwords>\"] [--by \u003Cname>] [--timeout \u003Cd> | --now [--yes]]\n                               pause the store: no pass starts new work — no step, try, round or\n                               check filed, no queued task admitted, no queued card started — and\n                               every session, job or attached, parks at its tool call's end: a call\n                               still in flight once the timeout has passed is cut. A check runs on, its\n                               end recorded; a landing whose check has ended commits or aborts.\n                               Pausing: N sessions finishing a tool call, cut in M min while they\n                               do, Paused once none does; nemia ls says each step paused.\n                               nemia start still writes a run, and approvals, retries and answers are\n                               kept, not acted on: the pass after nemia resume files what waits. nemia ls,\n                               nemia watch --once, nemia start and nemia up say it, and nemia ls names what finishes.\n                               Refused while a pause stands.\n                               --why: the reason, said wherever the pause is; --by: who pauses it\n                               (default: your OS user)\n                               --timeout: how long a call in flight is given to end, for this pause\n                               (90s, 10m, 2h; from 1s to 23h); default: the store's, nemia pause-timeout\n                               --now: pause at once, the board's Stop now — each step session at work,\n                               but a landing's, has its call cut at once and parks, going on as the\n                               same try on nemia resume;\n                               each check is asked to stop and run again. A pause that stands is kept.\n                               --yes: with --now, act; without it, --now only says what it would\n                               cut\n                               on the port: pause\n  nemia resume\n                               resume the store: each session parked goes on as the same try, and the\n                               next pass files what waits\n                               on the port: resume\n  nemia subscribe \u003Cname> \u003Ctask> [-l \u003Ckey>=\u003Cvalue>]... [--digest-every \u003Cminutes>] [--at-least \u003Clevel>]\n                               give the engine's long-lived task \u003Ctask> a subscription to the feed,\n                               named \u003Cname>, from its end: a line one of its selectors takes is said to\n                               it at once, and the rest folded into a digest said with the next wake,\n                               or after --digest-every minutes (default 60) with none.\n                               -l: the lines it takes, given again for another (default:\n                               waits=person, the lines that wait on a person). Refused for a task not\n                               long-lived.\n                               --at-least: only the lines of runs whose factory's level is at least\n                               this (manual, semi-auto, auto), read as each line is delivered.\n                               on the port: subscribe\n  nemia unsubscribe \u003Cname>\n                               take the subscription \u003Cname> away: its task is told nothing more of the feed\n                               on the port: unsubscribe\n  nemia policy propose \u003Cagent> --file \u003Cpath>|-\n                               the agent's proposal: rules from your custom instruction, as JSON,\n                               with the words the machinery cannot hold kept as guidance; none is in\n                               force until you confirm it (nemia policy confirm)\n                               on the port: policy_propose\n  nemia daemon [--json]\n                               the store's daemon, started where none ticks it (as any command that\n                               ticks starts it, unless NEMIA_DAEMON_START=off): its state, its pid and\n                               the url the api serves at, from its lease. --json: the whole answer\n                               on the port: daemon\n```\n\n## Prompts the person\n\nEach asks the person at a prompt every time it runs: say what it does before you run it.\n\n```text\n  nemia inbox rm \u003Cname>\n                               drop a card without starting it\n                               on the port: inbox_rm\n  nemia factory add \u003Cname> --repo \u003Cpath|url> [--repo \u003Cpath|url> ...] [--no-trust]\n                               file a factory: a lowercase slug naming one or more directories or git URLs,\n                               each URL cloned, each trusted for its steps unless --no-trust\n                               on the port: factory_add\n  nemia factory apply \u003Cname> --template \u003Cname|folder|archive> --repo \u003Cpath|url> [--repo \u003Cpath|url> ...] [--set \u003Ckey>=\u003Cvalue> ...] [--no-trust]\n                               make a factory from a template that passes its check, its parameters\n                               given with --set; the daemon renders its tree\n                               on the port: factory_apply\n  nemia factory adopt \u003Cname> --template \u003Cname|folder|archive> [--set \u003Ckey>=\u003Cvalue> ...] --plan \u003Cdigest>\n                               take a hand-made factory over by a template, in place, as nemia factory\n                               diff previewed it: its name, cards, runs and stats stay\n                               on the port: factory_adopt\n  nemia factory upgrade \u003Cname> --to \u003Cx.y.z> [--template \u003Cname|folder|archive>] [--set \u003Ckey>=\u003Cvalue> ...] [--plan \u003Cdigest>] [--downgrade]\n                               upgrade a templated factory: its file wants --to, and the daemon merges\n                               the template's change into its tree once each conflict is resolved;\n                               with --to the version applied and --set, a change of its parameters\n                               alone, filed with the --plan nemia factory diff printed for it\n                               on the port: factory_upgrade\n  nemia factory resolve \u003Cname> \u003Cconflict> --ours|--theirs|--file \u003Cpath>\n                               choose for one conflict of a templated factory's upgrade or switch: your\n                               tree, the template (none: drop the file), or the text of a file; once\n                               each conflict of a workflow, an agent, a skill, a hire or the rules and\n                               instruction has one, that object is proposed as they compose it, to\n                               approve (nemia proposals approve \u003Cp> -n \u003Cname>); a carried file stays\n                               a choice; the daemon merges once each has one and each proposal applied\n                               on the port: factory_resolve\n  nemia factory attach \u003Cfactory> \u003Cworkflow> [--repo \u003Cname>]\n                               offer a workflow in a factory, by its folder's name: a factory offers\n                               only the workflows attached to it; --repo names the repository its runs\n                               work on, by folder name, required where the factory has several\n                               on the port: factory_attach\n  nemia factory detach \u003Cfactory> \u003Cworkflow>\n                               take a workflow back, once no hire fills its steps\n                               on the port: factory_detach\n  nemia factory hire \u003Cfactory> \u003Cagent> --step \u003Cworkflow>\u002F\u003Cstep>\n                               hire an agent into a workflow's step: one a step, a second replaces it\n                               on the port: factory_hire\n  nemia factory fire \u003Cfactory> --step \u003Cworkflow>\u002F\u003Cstep>\n                               take a step's hire back\n                               on the port: factory_fire\n  nemia factory schedule \u003Cfactory> \u003Cworkflow> --cron \u003Cexpr> --brief \u003Ctext> [--tz \u003Czone>]\n                               start a run of an attached workflow at each fire of a cron, by the\n                               daemon, at every level; a second replaces the first\n                               on the port: factory_schedule\n  nemia factory unschedule \u003Cfactory> \u003Cworkflow>\n                               take a workflow's schedule back: the daemon starts it no more\n                               on the port: factory_unschedule\n  nemia factory server add \u003Cfactory> \u003Cserver> --url \u003Caddress> [--transport http|sse] [--header \u003Cname>=\u003Cvalue> ...] [--client-id \u003Cid>] [--callback-port \u003Cport>]\n                               record a remote MCP server for a factory: its steps' sessions are\n                               handed it, and an agent `mcp:` line names it\n                               on the port: factory_server_add\n  nemia factory server rm \u003Cfactory> \u003Cserver>\n                               take a server off a factory's record\n                               on the port: factory_server_rm\n  nemia factory volume add \u003Cfactory> \u003Cpath>:ro|:rw\n                               mount a folder of this machine for every local step of a factory's runs,\n                               read-only (:ro) or writable (:rw); a writable one is not a run's workspace\n                               on the port: factory_volume_add\n  nemia factory volume rm \u003Cfactory> \u003Cpath>\n                               take a volume off a factory, by its path\n                               on the port: factory_volume_rm\n  nemia factory setting set \u003Cfactory> \u003Charness>.\u003Ckey> \u003Cjson>\n                               give every session of a factory's runs on a harness one of its settings,\n                               claude.attribution among them; its value is JSON\n                               on the port: factory_setting_set\n  nemia factory setting rm \u003Cfactory> \u003Charness>.\u003Ckey>\n                               take a harness setting off a factory, by its key\n                               on the port: factory_setting_rm\n  nemia factory cost-line set \u003Cfactory> \u003Cworkflow> \u003Cusd>\n                               set what a run of a workflow may cost before its gate says so, in dollars:\n                               a gate whose caps pass it is said, and never approved unasked at auto\n                               on the port: factory_cost-line_set\n  nemia factory cost-line rm \u003Cfactory> \u003Cworkflow>\n                               take a workflow's cost line off a factory\n                               on the port: factory_cost-line_rm\n  nemia factory instruct \u003Cname> --file \u003Cpath>|- | --clear\n                               write the instruction the store keeps for a factory (- for stdin), what\n                               Nemia is to know of it and no step is told, or remove it with --clear\n                               on the port: factory_instruct, factory_instruct_delete\n  nemia factory revisions \u003Cname> follow|pinned\n                               whether a factory's runs follow merged revisions at each filing (follow)\n                               or keep them until nemia refresh (pinned, the default)\n                               on the port: factory_revisions\n  nemia factory trust \u003Cname>\n                               trust a factory's places for the harnesses its steps run on, so each\n                               place's own settings apply to them\n                               on the port: factory_trust\n  nemia factory rm \u003Cname>\n                               remove a factory's file, and the instruction and the rules the store\n                               keeps for it\n                               on the port: factory_rm\n  nemia template export \u003Cfolder> --from \u003Crepo> [--factory \u003Cname>]\n                               write a template from a repository and a factory, and say where each\n                               part came from\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia template pack \u003Cfolder> [--out \u003Cfile>]\n                               pack a template's folder that passes its check into one archive, and print\n                               the digest of its content\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia template unpack \u003Carchive> \u003Cdir>\n                               write the folder a template's archive holds at \u003Cdir>\u002F\u003Cname>, each file's\n                               bytes and mode as packed\n                               CLI only: a template is a folder on the machine nemia runs on; the board reads none yet\n  nemia agents write (yours | factory) [\u003Cname>] --file \u003Cname>.md [--factory \u003Cname>] [--version \u003Cv>]\n                               an agent of yours, or of a templated --factory, written from its file:\n                               \u003Cname> is the agent it changes or renames, at the --version it was\n                               read at; none, a new one\n                               on the port: agents_write, agents_replace\n  nemia agents copy \u003Cfile> [--as \u003Cname>] [--keep \u003Cpart>]...\n                               an agent file copied into your library, with its schema and skills;\n                               a part your library holds otherwise stops it: --as or --keep\n                               on the port: agents_copy\n  nemia agents rm yours \u003Cname> --version \u003Cv>\n                               an agent of yours removed, at the --version it was read at\n                               on the port: agents_rm\n  nemia skills write (yours | factory) [\u003Cname>] --file \u003CSKILL.md> [--factory \u003Cname>] [--version \u003Cv>]\n                               a skill of yours, or of a templated --factory, written from its\n                               SKILL.md: \u003Cname> is the skill it changes, at the --version it was\n                               read at; none, a new one\n                               on the port: skills_write, skills_replace, skills_copy\n  nemia cancel \u003CrunId>\n                               stop what it has in flight; file nothing more\n                               on the port: cancel\n  nemia rm \u003CrunId> [\u003CrunId> ...] [--delete-branches [--force-unmerged]]\n                               forget runs that are over, a line each: a run's worktrees removed\n                               through git, what one held uncommitted committed first, then its\n                               record and every file under WF_ROOT keyed by its id. Refused while\n                               a task of it is in flight, anything is due to be filed or asked\n                               for, or a run waiting on it has not started. Its branches are kept:\n                               --delete-branches deletes those merged into the run's base branch,\n                               and --force-unmerged beside it the others, with their work. Then\n                               the engine forgets the run's tasks: their journals and session\n                               files; one not settled yet is forgotten by the next passes\n                               on the port: rm\n  nemia prune [\u003CrunId> ...] [--unmerged] [--artifacts] [--yes]\n                               free the space runs that are over hold: their worktrees, removed\n                               through git, and nothing else — each run's record, branches and\n                               files are kept, and it stays on nemia ls, nemia decide and the board,\n                               marked pruned. --artifacts removes its files too (runs\u002F\u003Cid>\u002F\n                               artifacts\u002F: screenshots, logs), of a run whose worktrees are\n                               already gone as well; its record and branches are still kept.\n                               Lists what it would remove, with its size, and removes it on\n                               --yes, or on y at the prompt. With no run, every run. Refused\n                               while a run is not over, owes a hand-off, is a folder run, or has\n                               a worktree held, on no branch, or with changes no commit has;\n                               and a finished run not merged unless --unmerged, with --artifacts\n                               as without: its files are what a person reads to decide it\n                               on the port: prune\n  nemia pause-timeout [\u003Cd>]\n                               the store's pause timeout: how long nemia pause, nemia service restart and\n                               nemia upgrade give each tool call in flight to end before it is cut; bare,\n                               the timeout in force (10m by default); given one, sets it: digits and s, m\n                               or h, from 1s to 23h, as 90s, 10m or 2h. nemia pause --timeout \u003Cd> gives one\n                               pause a timeout of its own\n                               on the port: pause-timeout, pause_timeout_set\n  nemia policy confirm \u003Cagent> \u003Cid> --words '\u003Cwords>'\n                               put in force a rule the agent proposed, or the default again, as its\n                               words read: from then on, at auto, the answer it names is given unasked\n                               where it holds. Refused once the rule reads otherwise. Your first confirm\n                               keeps the default beside it. Asked of you every time\n                               on the port: policy_confirm\n  nemia policy rm \u003Cagent> \u003Cid>\n                               take a rule out of force, the default among them: what it allowed is\n                               asked of you again. Asked of you every time\n                               on the port: policy_rm\n  nemia service install [--env KEY=VALUE ...] [--size auto] [--bind \u003Caddr> [--public]] [--name \u003Cname>] [--take-over \u003Cpid>]\n                               install the daemon as a login service: a systemd user unit on Linux,\n                               a launchd agent on macOS. It runs the api of this checkout with this\n                               node, WF_ROOT and PATH, and each --env, written once; it restarts on\n                               failure and starts at login. Started now unless another daemon holds\n                               the store; --take-over \u003Cpid> pauses the store, waits for the calls\n                               in flight of its sessions, stops that pid and starts the service\n                               (checks and jobs run on, as nemia service restart --help says).\n                               --size auto: the sessions at once and each one's memory limit and\n                               request, read off this machine (WF_MAX_CONCURRENT, WF_MEMORY_LIMIT,\n                               WF_MEMORY_REQUEST; neither memory one on macOS), each but one --env\n                               gives; the floor kept, WF_MEMORY_RESERVED, is its default (an eighth\n                               of the machine, 1 GiB at least). On systemd it also writes\n                               \u003Cname>-sessions.slice, its MemoryMax at the machine less that floor,\n                               and points WF_CGROUP_PARENT at it, so the sessions together are\n                               held to what the machine allocates (none written where --env gives\n                               WF_CGROUP_PARENT: a line names its set-property). --bind \u003Caddr>: the\n                               address the daemon listens on (NEMIA_BIND), and the store marked\n                               exposed (WF_EXPOSED=1), so every caller presents a token: nemia\n                               pairings approve and nemia board link hand out the first. A public or\n                               wildcard address is refused unless --public acknowledges it\n                               (NEMIA_BIND_PUBLIC=1); nemia service addresses lists this machine's.\n                               --name: the service name (nemia; required for a store that is not\n                               the default)\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service uninstall [--name \u003Cname>]\n                               stop the login service, disable it and remove its file and the\n                               sessions' slice --size auto wrote; a daemon started by hand is left\n                               alone\n                               CLI only: it writes or reads the person's login services, on the machine the command runs on\n  nemia service restart [--name \u003Cname>] [--timeout \u003Cd>]\n                               restart the service's daemon on the code it runs: pause the store, give\n                               the daemon's sessions up to the timeout to end their tool calls\n                               (--timeout \u003Cd>, else nemia pause-timeout, 10m by default), stop it (up to\n                               90s, and until the unit has ended its process), start the service (up to\n                               60s), and resume once the new daemon has completed its first pass;\n                               refused, naming the log, when that pass fails, and naming the command\n                               that starts the service when it is not running.\n                               Jobs and checks, a landing waiting on CI among them, run on and are not\n                               waited for; each session parks and resumes as the same try. Each wait is\n                               said on stderr as it waits. A pause a person put stands; the restart's\n                               own ends with the command: killed, the next pass of the daemon lifts it.\n                               Starts the service where nothing ticks; a daemon started by hand is\n                               named with --take-over's line\n                               CLI only: it restarts the person's login service, on the machine the command runs on\n  nemia upgrade [\u003Ctag>] [--repo \u003Cpath>] [--name \u003Cname>] [--timeout \u003Cd>]\n                               on an archive install, download the tag's archive (the newest where\n                               none is named), check it against its SHA256SUMS and run its nemia\n                               version, into the store's releases\u002F\u003Ctag>; on a release built from a\n                               repository, build the tag from a clean export into releases\u002F\u003Ctag>:\n                               npm ci, npm run build and npm run release:check there, on a trial\n                               store; then point releases\u002Fcurrent at it and restart the daemon onto\n                               it where the login service runs the pointer: jobs run on, and the\n                               daemon's sessions are given up to the timeout to end their tool calls\n                               (--timeout \u003Cd>, else nemia pause-timeout), park, and resume as the same try on\n                               the new release; then say each piece the new release reads missing\n                               that a run waits on, and each of the machine and its service, with\n                               its fix, or that none of those is. A tag built already is switched\n                               to without a rebuild, so an older tag rolls back. --repo:\n                               the repository the tag is in (the running release's source);\n                               --name: the service (nemia). Then it removes each release that is\n                               not current, not among the two newest, and that no session or\n                               running nemia runs on, the folder of each build whose upgrade died,\n                               and each build log older than 30 days, as the daemon does every five\n                               minutes. A tag older than the oldest release that can read the store\n                               (store.json) is refused before anything is fetched or built. It\n                               upgrades this machine's install, whatever the context: the server a\n                               context names is upgraded from here with nemia server upgrade [\u003Ctag>],\n                               through its api (not nemia service, this machine's login unit)\n                               on the port: upgrade\n  nemia frontmatter rewrite [\u003Cfolder>...] [--check]\n                               rewrite as YAML each markdown file's frontmatter still in the former\n                               line form, under the folders given or the store's own. --check: write\n                               nothing, and exit 1 where a rewrite is due\n                               CLI only: it rewrites files on the machine nemia runs on, folders a person names among them\n  nemia workflows write (yours | factory) \u003Cname> --folder \u003Cdraft> [--factory \u003Cname>] [--version \u003Cv>]\n                               a workflow folder of yours, or of a templated --factory, written from\n                               a draft folder, its steps, README, scripts and schemas, leaving a file\n                               it does not carry; one there already, at the --version it was read at\n                               on the port: workflows_write, workflows_replace\n  nemia workflows rm yours \u003Cname> --version \u003Cv>\n                               a workflow folder of yours removed, at the --version it was read at\n                               on the port: workflows_rm\n```\n\n## Answers what waits\n\nEach answers what waits on the person. Bring the matter first, with the `bring-what-waits` skill, then run the verb they choose, at a shell with their words as `--message`, or its tool on the port with them as its message: its prompt is their decision, asked once. `retry` asks only of a step that waits on the person, or with `--budget`.\n\n```text\n  nemia retry \u003CrunId> [\u003Cstep>] [--continue | --fresh | --model \u003Cm> | --recheck | --round] [--answer \u003Cn>=\u003Cletter>]... [--message \"\u003Canswer>\"] [--budget \u003Camount>]\n                               try a step waiting on you again, once, on the next tick:\n                               --continue (default) continues its session, told what happened;\n                               --fresh a new session on its branch; --model \u003Cm> the same, on m;\n                               --recheck runs its check again, and files no session;\n                               --round: once more — a loop out of rounds goes round again, once,\n                               the step it loops to continued, then a new reviewer; the step is the\n                               one that closes the loop, left out where the run has one loop;\n                               --message \"\u003Canswer>\" goes into the try's brief, as ## Your answer:\n                               for a step that signed off blocked, the answer to its needs — or\n                               what you have to say to any step you try again; not with --recheck\n                               or --round. The next argument is the answer, whatever it starts with;\n                               an answer that is only one of these flags' names goes as --message=\u003Canswer>;\n                               --budget \u003Camount>: for a step with a budget that waits on you, its\n                               cap raised to that many dollars, above its cap and what it has cost;\n                               with --round, the cap of a step of the loop over its budget;\n                               --answer \u003Cn>=\u003Cletter>: answer question n, numbered as nemia decide\n                               \u003CrunId> prints them, with the option its letter names; once for each\n                               question: \"\u003Cn>=a,c\" chooses several where it takes several,\n                               \"\u003Cn>=a: \u003Cwords>\" adds your own words, \"\u003Cn>: \u003Cwords>\" says words\n                               alone; --message is then the note under your answers\n                               on the port: retry\n  nemia approve \u003CrunId> [\u003Cstep>] [--answer \u003Cn>=\u003Cletter>]... [--budget \u003Cstep>=\u003Camount>]... [--message \"\u003Cwords>\"] | \u003CrunId> \u003Cstep> --off\n                               let a step waiting at its gate hand on: what waits on it is filed on\n                               the next tick, and every step after the gate is handed your words\n                               under ## Approved. To send it back instead, with what to change:\n                               nemia retry \u003CrunId> \u003Cstep> --message \"\u003Cobjection>\"\n                               Of a loop out of rounds — the step that closes it, or the run alone\n                               where it has one loop — enough: the loop ends on your words, with its\n                               findings open, --message \"\u003Cwhy>\" required. What waits on the looping\n                               step is filed as if it had said again: false, and every step after\n                               the loop is handed your words and those findings, under the heading\n                               of a loop ended as it stands.\n                               To go round once more instead: nemia retry \u003CrunId> [\u003Cstep>] --round\n                               --answer \u003Cn>=\u003Cletter>: answer question n, numbered as nemia decide\n                               \u003CrunId> prints them, with the option its letter names; once for each\n                               question: \"\u003Cn>=a,c\" chooses several where it takes several,\n                               \"\u003Cn>=a: \u003Cwords>\" adds your own words, \"\u003Cn>: \u003Cwords>\" says words\n                               alone; --message is then the note under your answers\n                               --budget \u003Cstep>=\u003Camount>: the cap of a step after the gate, in\n                               dollars, over what the step proposed for it; once for each step.\n                               Your approval makes the budgets the step proposed the caps of the\n                               steps after the gate, and auto never approves unasked a cap above\n                               a step's own budget\n                               --off: take back the approval of a step that stands approved while\n                               nothing it approved has been filed — an ending waiting on the base\n                               branch, held, or not yet ticked. The step waits at its gate again\n                               on the port: approve\n```\n",{"name":5,"description":7},"main","e3efae8a132cdd1bb19337ef25a64fc7444f2708","packages\u002Fagent-orchestrator\u002Fnemia\u002Fskills\u002Fverbs",0,"https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia","https:\u002F\u002Fgithub.com\u002Femyann","Senior Software Engineer","2026-11-15 05:55:45",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add emyann\u002Fnemia --skill verbs",{"tool":38,"command":39},"bunx","bunx skills add emyann\u002Fnemia --skill verbs",{"tool":41,"command":42},"pnpm","pnpm dlx skills add emyann\u002Fnemia --skill verbs","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Fagent-orchestrator\u002Fnemia\u002Fskills\u002Fverbs\u002FSKILL.md",true,[],[47,54,61,68,75,82],{"skillNo":48,"name":49,"nameZh":6,"description":50,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7BGEY7EKSRDFAB7B9567QKJ","groom-the-inbox","Read each card in play in the inbox against the repository and say, per card, done (the commit on the base that landed it), replaced (the decision or card that did), stale (each fact that moved, what it said and what it says now) or standing, reading closely only what the period touched. For the janitor's step, which proposes and never acts: a person removes or amends each card by the command the check writes.","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Fgroom-the-inbox\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7B4AD0ZJYZ3HAJCCBM4CPTY","write-a-spec","Turn an accepted intent into a spec an engineer can build from in this repository: the goal restated, the approach and why, every path it touches, the steps, the tests that prove it, the risks, and what is still open. For a step that designs a change before anyone builds it.","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Fwrite-a-spec\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7B4A88P2FKWCS1RND507WHE","review-a-change","Review a change in this repository as a second engineer would: against its intent and spec, by reading the diff and the suite's verdict, with findings that name the file, the line and what goes wrong. For a read-only step that decides whether the work goes round again.","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Freview-a-change\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M7B4A2HFW8F53E390A2KM0R8","read-the-stats","Read a factory's stats since the janitor's last run, against the same length of time before it, write a short report, and propose one correction per card — an allow or deny by name, a line in an agent's body or a skill, or a refusal kept on purpose — each with the numbers behind it, never one already in the inbox. For the janitor's step, which proposes and never acts.","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Fread-the-stats\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7B49Z4GJ2G8T76X744YHAAN","land-a-branch","Land a run's result on the base branch in this repository's own checkout, the way the step says it lands: make that one act, read its result, and write the message that records what landed — or change nothing and say why it refused. For a gated ending step (workspace: base).","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Fland-a-branch\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7B49TMM0QAWF5WQ0QBXM9BG","implement-a-change","Build a change from its spec in this repository: tests first where a rule is claimed, small commits, the tests of what changed green for the right reason, and each review finding answered in the next round. For a step that writes the code.","https:\u002F\u002Fgithub.com\u002Femyann\u002Fnemia\u002Fblob\u002Fmain\u002Fpackages\u002Ftemplates\u002Fsdlc\u002Fskills\u002Fimplement-a-change\u002FSKILL.md",[],[]]