[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6F5DFM61CM217R296RS9CH3":3,"skill-same:01M6F5DFM61CM217R296RS9CH3":84},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":41,"linkOnly":42,"markdownContent":43,"markdownBody":44,"frontMatter":45,"refName":53,"commitSha":54,"skillDirectory":55,"forksCount":56,"repositoryHtmlUrl":57,"repositoryHomepageUrl":6,"ownerHtmlUrl":58,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":59,"install":60,"resources":73},"01M6F5DFM61CM217R296RS9CH3","delegate-setup",null,"Configure approved delegation lanes across installed implementer CLIs, including optional model and effort choices, then write global or project config only after explicit user approval.","boisenoise\u002Fskills-collections","boisenoise","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F193180?v=4","01M6CDBDQK6BDN55ZMA6B6PKBM","01M6CDBDVSJ3JPTBYN26FZ7TZK",12,"MIT","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-delegate-setup\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40],"setup","delegate setup","delegate","2026","safe","blob","agent-orchestration","project","across","community","master","only","implementer","write","node","model","configure","https","license","delegation","config","orchestration","github","2026-11-10 10:12:30",false,"---\nname: delegate-setup\ndescription: Configure approved delegation lanes across installed implementer CLIs,\n  including optional model and effort choices, then write global or project config\n  only after explicit user approval.\nrisk: safe\ncategory: agent-orchestration\nsource: https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\nsource_repo: amElnagdy\u002Fdelegate-skills\nsource_type: community\ndate_added: '2026-08-26'\nlicense: MIT\nlicense_source: https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\u002Fblob\u002Fmaster\u002FLICENSE\ncompatibility: Requires Node 18+. No implementer CLIs are required — the skill discovers\n  what is available.\nmetadata:\n  version: 0.5.0\n---\n# Delegate Setup\n\n## When to Use\n\n- You want to configure which implementer CLI handles which kind of work (fleet lanes).\n- You need to discover installed implementers and write lane config after user approval.\n\nYou are the **orchestrator** in **setup mode**. Discover installed implementer CLIs, propose a\n**fleet of lanes**, and write configuration only after the user approves.\n\nThis skill does **not** dispatch coding work. It only authors the lane map.\n\nOne concept: **lanes**. Never say “routes.”\n\nExample lane: **feature** → implementer `opencode`, model `opencode\u002Fgrok`, variant `high`\n(OpenCode uses `variant` for reasoning intensity, not `effort`).\n\n## When NOT to use this\n\n- The user wants a task implemented — use the matching `*-delegate` skill instead.\n- A one-off model change on a single dispatch — pass `--model` \u002F `--effort` \u002F `--variant` on that relay.\n\n## Hard rules\n\n1. Every lane **must** include `implementer`.\n2. Put dials on the same object (`model`, `effort` or `variant`, …) only if that implementer supports them — see [references\u002Fschema.md](references\u002Fschema.md).\n3. Show a human-readable lane table **and** the full JSON before every write; re-show after every tweak.\n4. Write **only** after an explicit approval (“yes”, “approve”, “write it”).\n5. Ask scope unless already clear: **global** (all projects) vs **this repo only**. Never create a project file just because cwd is a git repo. If there is no git repo, default to global and say so.\n6. Do not invent model identifiers.\n7. In interview or usage-scan mode, never write **any** dial the user did not give you and the schema does not require — omit it, so the CLI’s or relay’s own default applies.\n8. Prefer 3–5 useful lanes over a kitchen-sink map.\n9. Never edit `AGENTS.md`, `CLAUDE.md`, or other user agent-instruction files.\n10. Never run a `*-delegate` relay from this skill.\n\n(`\u003Cskill-dir>` is this skill’s install directory — the folder that contains this `SKILL.md`.)\n\n## Flow\n\n`discover → load → grounding menu → propose (with Basis) → scope → approve → write`\n\n### 1. Verify runtime availability, then discover\n\nThis is a docs-only tree: scripts\u002Fdiscover.mjs, config.mjs and lane.mjs are absent. Do not run the example commands against nonexistent paths or fabricate discovery\u002Ftrust results. If a separately installed runtime exists, inspect its verified revision, license and behavior before using it. Otherwise produce the proposed table\u002FJSON only, clearly marked unvalidated, and stop before config writes that depend on the missing validator\u002Fapproval-hash mechanism.\n\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fdiscover.mjs\"\n```\n\nSummarize installed vs missing, auth (`true` \u002F `false` \u002F `null` = unknown), and whether models were\n`reported`, `aliases` (curated aliases in the registry, not live discovery — full model names also\nwork), `unsupported`, or `failed`.\n\n### 2. Load existing (effective map)\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" load --cwd \"$PWD\"\n```\n\n- Neither present → “No lanes configured yet.”\n- Otherwise → table of **effective** lanes with a Source column (`global` \u002F `project`). Do not paste\n  both raw files unless asked.\n- If `projectPresent` is true and `projectTrusted` is false, label the project lanes **untrusted**.\n  They cannot dispatch until the user reviews and approves a project write.\n\n### 3. Propose\n\nDiscovery reports capability, never task fit. So ask **one** grounding question before proposing\nanything — one question, three options, not a wizard:\n\n> How should I pick the lanes? **(1) Quick defaults** — I decide, no questions.\n> **(2) Interview** — about four questions on how you want work allocated.\n> **(3) Usage scan** — I re-read your CLIs’ local session folders (counts and dates only, never the\n> conversations) and let the numbers place your lanes — if one CLI dominates, expect one question\n> about its role. Happy to do 2 and 3 together.\n\n- **Quick defaults** → propose immediately.\n- **Interview** → the four questions (allocation policy, never model rankings) and how to ask them\n  (one medium per round) live in [references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md) — read\n  it before you ask.\n- **Usage scan** → `node \"\u003Cskill-dir>\u002Fscripts\u002Fdiscover.mjs\" --usage`. Tell the user it is metadata\n  only before running it. Each discovered CLI gains `usage: { sessions, lastUsed }`; `null` means no\n  probe is wired — unknown, not unused.\n- **Both** → run the scan first, then ask only what the numbers cannot answer.\n- Inside a git repo, repo signals (languages, test weight, frontend share) are a fourth source of\n  evidence. They do not change the menu; they feed the proposal and the `repo` basis.\n\n**That menu is also the consent surface** — the option chosen sets how much of the map is yours to\ndecide:\n\n- **Quick defaults** — the user hired your opinion. A full map is legitimate, dials included; label\n  every lane `my opinion`, say plainly that the map is your opinion, and keep it cheap to revise.\n- **Interview \u002F usage scan** — evidence modes, so **every** dial is gated (rule 7): set one only from\n  the user’s answer, or where the schema requires it (opencode lanes require `model`). Omitting is\n  always safe — every dial has a default the user already lives with, and a CLI’s configured default\n  is their standing choice, better evidence than your priors. Choosing which installed implementer\n  gets a lane is still yours — Basis `my opinion` — but a dial that raises spend is not: offer your\n  dial picks only as an addendum after the proposal, see\n  [references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n- **An unanswered question shrinks the map; it never licenses a substitution.** Propose fewer, more\n  conservative lanes, name the axis you are blind on (no quota answer → say the map is quota-blind),\n  and invite the answer anytime. Re-ask once at most; never backfill silence with priors.\n\n**Delegation economics.** The orchestrator reviews and lands every result — the review is the\nquality gate, so optimize total cost, not implementer prestige:\n\n- Prefer capable, authenticated, burnable, **low-usage** CLIs for bounded, objectively gated work\n  (tests, mechanical refactors, straightforward fixes) when their reliability keeps review and\n  rework economical — lanes push token burn away from the subscriptions the user is protecting.\n  Low usage alone does not establish burnable: discovery cannot see plans, limits, or per-run\n  cost, and a rarely-used CLI may be metered or deliberately avoided. Burnable comes from the\n  user's quota answer — or, in quick defaults, from your labeled opinion.\n- Avoid binding a lane to a CLI the user is protecting or orchestrates from, by default; bind it\n  only when the user asks for it or no acceptable alternative exists. Lanes are\n  **orchestrator-blind**: the same lane fires from every seat the user drives from, and from that\n  CLI's own seat it dispatches the CLI to itself.\n- Surplus placement breaks down when rework and review cost exceed the savings; when the\n  implementer is flaky; when correctness rides on security, concurrency, migrations, or unstated\n  domain knowledge; and when the output **is** the product (debate, architecture, research) —\n  review limits damage, it does not manufacture a good first attempt. Bind those lanes to stronger\n  implementers.\n- An explicit \"spare X\" answer removes X from proposed lanes by default, and overrides blanket\n  posture answers on any lane the user explicitly retains for X — ask whether the posture applies\n  there; omit the dial if unanswered. Never silently stretch one answer across an axis it\n  conflicts with.\n\nQuestion phrasings for the burn\u002Fspare and trust interview live in\n[references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n\nThen propose the lanes. Name them after the work the user described; fall back to `feature`, `tests`,\n`ui`, `fast`, `complex`. Installed implementers only.\n\nShow:\n\n| Lane | Implementer | Model | Effort \u002F variant | Basis | Source (if updating) |\n| --- | --- | --- | --- | --- | --- |\n| feature | opencode | opencode\u002Fgrok | variant: high | your answer + schema requirement | — |\n| tests | codex | — | — | usage data | — |\n| ui | claude | — | — | my opinion (implementer) | — |\n\n**Basis** is mandatory on every lane: `your answer` \u002F `usage data` \u002F `repo` \u002F `my opinion` \u002F\n`schema requirement` (a dial the schema forces is neither evidence nor opinion — say so). A lane you\npicked from model-quality priors is `my opinion` — never present it as something the tooling\ndetermined, and “installed and authenticated” is capability, not evidence of fit. When a lane’s\nimplementer and its dials come from different places, split the label — see\n[references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n\nThen the **complete** JSON (`version`: `delegate-fleet.v1`). One line of why per lane; flag auth or\nmodel uncertainty.\n\nSchema and dial table: [references\u002Fschema.md](references\u002Fschema.md).\n\n### 4. Scope\n\n- User said global \u002F all projects \u002F outside the project → `global`.\n- No git repo → `global` (say so).\n- Else ask once: global vs this repo only.\n\n### 5. Approve and write\n\nOn explicit yes, write **only** the chosen scope (validate first). Build the payload from that\nscope’s raw file (or an empty `lanes` object if new) — not from the effective merged `load` view,\nor a project write will shadow global-only lanes and a global write will promote project-only ones.\n\nCreate a uniquely named file under the platform temporary directory (`$TMPDIR`, `%TEMP%`, or Node\n`os.tmpdir()`; never hard-code `\u002Ftmp`, which breaks on native Windows), write the **exact approved\nJSON** into it with the orchestrator's file-writing tool, and use that populated path as\n`\u003Clanes-json>` below. Never validate an empty temp file. Remove the temp file after the\nvalidation\u002Fwrite attempt, whether it succeeds or fails.\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" validate \"\u003Clanes-json>\"\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" write --scope global \"\u003Clanes-json>\"\n# or:  write --scope project --cwd \u002Fpath\u002Fto\u002Frepo \"\u003Clanes-json>\"\n```\n\nRe-read with `load`, then confirm the path written and the active lane names. Project writes bind\napproval to the exact config content; later changes fail closed until re-approved. On update, a short\nbefore\u002Fafter is enough.\n\n### 6. Ready to delegate\n\nStop after confirming. Tell the user the map is ready. For later work: read the lane’s\n`implementer`, load that `*-delegate` skill, and dispatch with `--lane \u003Cname>` (explicit\n`--model` \u002F `--effort` \u002F `--variant` still win when passed). Do not start a delegate task\nunless they ask.\n\n## Reconfigure\n\nSame flow. Show the effective current map, propose changes, approve, write one scope’s file.\nReinstalling the skills package must not rewrite these files — they live outside the package.\n\n\n## Limitations\n\n- Never dispatches work itself — only discovers CLIs and writes config after explicit approval.\n- Docs-only import — executable helpers (`scripts\u002F`) not included; see upstream for full runtime. Requires Node 18+.\n\n> Adapted from [amElnagdy\u002Fdelegate-skills](https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills) (MIT) — docs-only, runtime not bundled.\n\n## Configuration trust limits\n\nThe schema reference describes upstream behavior, not an implementation supplied here. Installed\u002Fauthenticated does not establish supported model IDs, quota, safe defaults or enforced read-only behavior. A missing runtime means no project trust hash can be claimed. Before an approved write, inspect symlink destinations, preserve unrelated lanes\u002Fsettings, use private temporary files and avoid exposing session filenames\u002Faccount metadata. A usage scan remains opt-in and requires verifying its actual implementation reads only the promised metadata. No discovery, session scan, config write or delegation occurred during this review.\n","# Delegate Setup\n\n## When to Use\n\n- You want to configure which implementer CLI handles which kind of work (fleet lanes).\n- You need to discover installed implementers and write lane config after user approval.\n\nYou are the **orchestrator** in **setup mode**. Discover installed implementer CLIs, propose a\n**fleet of lanes**, and write configuration only after the user approves.\n\nThis skill does **not** dispatch coding work. It only authors the lane map.\n\nOne concept: **lanes**. Never say “routes.”\n\nExample lane: **feature** → implementer `opencode`, model `opencode\u002Fgrok`, variant `high`\n(OpenCode uses `variant` for reasoning intensity, not `effort`).\n\n## When NOT to use this\n\n- The user wants a task implemented — use the matching `*-delegate` skill instead.\n- A one-off model change on a single dispatch — pass `--model` \u002F `--effort` \u002F `--variant` on that relay.\n\n## Hard rules\n\n1. Every lane **must** include `implementer`.\n2. Put dials on the same object (`model`, `effort` or `variant`, …) only if that implementer supports them — see [references\u002Fschema.md](references\u002Fschema.md).\n3. Show a human-readable lane table **and** the full JSON before every write; re-show after every tweak.\n4. Write **only** after an explicit approval (“yes”, “approve”, “write it”).\n5. Ask scope unless already clear: **global** (all projects) vs **this repo only**. Never create a project file just because cwd is a git repo. If there is no git repo, default to global and say so.\n6. Do not invent model identifiers.\n7. In interview or usage-scan mode, never write **any** dial the user did not give you and the schema does not require — omit it, so the CLI’s or relay’s own default applies.\n8. Prefer 3–5 useful lanes over a kitchen-sink map.\n9. Never edit `AGENTS.md`, `CLAUDE.md`, or other user agent-instruction files.\n10. Never run a `*-delegate` relay from this skill.\n\n(`\u003Cskill-dir>` is this skill’s install directory — the folder that contains this `SKILL.md`.)\n\n## Flow\n\n`discover → load → grounding menu → propose (with Basis) → scope → approve → write`\n\n### 1. Verify runtime availability, then discover\n\nThis is a docs-only tree: scripts\u002Fdiscover.mjs, config.mjs and lane.mjs are absent. Do not run the example commands against nonexistent paths or fabricate discovery\u002Ftrust results. If a separately installed runtime exists, inspect its verified revision, license and behavior before using it. Otherwise produce the proposed table\u002FJSON only, clearly marked unvalidated, and stop before config writes that depend on the missing validator\u002Fapproval-hash mechanism.\n\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fdiscover.mjs\"\n```\n\nSummarize installed vs missing, auth (`true` \u002F `false` \u002F `null` = unknown), and whether models were\n`reported`, `aliases` (curated aliases in the registry, not live discovery — full model names also\nwork), `unsupported`, or `failed`.\n\n### 2. Load existing (effective map)\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" load --cwd \"$PWD\"\n```\n\n- Neither present → “No lanes configured yet.”\n- Otherwise → table of **effective** lanes with a Source column (`global` \u002F `project`). Do not paste\n  both raw files unless asked.\n- If `projectPresent` is true and `projectTrusted` is false, label the project lanes **untrusted**.\n  They cannot dispatch until the user reviews and approves a project write.\n\n### 3. Propose\n\nDiscovery reports capability, never task fit. So ask **one** grounding question before proposing\nanything — one question, three options, not a wizard:\n\n> How should I pick the lanes? **(1) Quick defaults** — I decide, no questions.\n> **(2) Interview** — about four questions on how you want work allocated.\n> **(3) Usage scan** — I re-read your CLIs’ local session folders (counts and dates only, never the\n> conversations) and let the numbers place your lanes — if one CLI dominates, expect one question\n> about its role. Happy to do 2 and 3 together.\n\n- **Quick defaults** → propose immediately.\n- **Interview** → the four questions (allocation policy, never model rankings) and how to ask them\n  (one medium per round) live in [references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md) — read\n  it before you ask.\n- **Usage scan** → `node \"\u003Cskill-dir>\u002Fscripts\u002Fdiscover.mjs\" --usage`. Tell the user it is metadata\n  only before running it. Each discovered CLI gains `usage: { sessions, lastUsed }`; `null` means no\n  probe is wired — unknown, not unused.\n- **Both** → run the scan first, then ask only what the numbers cannot answer.\n- Inside a git repo, repo signals (languages, test weight, frontend share) are a fourth source of\n  evidence. They do not change the menu; they feed the proposal and the `repo` basis.\n\n**That menu is also the consent surface** — the option chosen sets how much of the map is yours to\ndecide:\n\n- **Quick defaults** — the user hired your opinion. A full map is legitimate, dials included; label\n  every lane `my opinion`, say plainly that the map is your opinion, and keep it cheap to revise.\n- **Interview \u002F usage scan** — evidence modes, so **every** dial is gated (rule 7): set one only from\n  the user’s answer, or where the schema requires it (opencode lanes require `model`). Omitting is\n  always safe — every dial has a default the user already lives with, and a CLI’s configured default\n  is their standing choice, better evidence than your priors. Choosing which installed implementer\n  gets a lane is still yours — Basis `my opinion` — but a dial that raises spend is not: offer your\n  dial picks only as an addendum after the proposal, see\n  [references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n- **An unanswered question shrinks the map; it never licenses a substitution.** Propose fewer, more\n  conservative lanes, name the axis you are blind on (no quota answer → say the map is quota-blind),\n  and invite the answer anytime. Re-ask once at most; never backfill silence with priors.\n\n**Delegation economics.** The orchestrator reviews and lands every result — the review is the\nquality gate, so optimize total cost, not implementer prestige:\n\n- Prefer capable, authenticated, burnable, **low-usage** CLIs for bounded, objectively gated work\n  (tests, mechanical refactors, straightforward fixes) when their reliability keeps review and\n  rework economical — lanes push token burn away from the subscriptions the user is protecting.\n  Low usage alone does not establish burnable: discovery cannot see plans, limits, or per-run\n  cost, and a rarely-used CLI may be metered or deliberately avoided. Burnable comes from the\n  user's quota answer — or, in quick defaults, from your labeled opinion.\n- Avoid binding a lane to a CLI the user is protecting or orchestrates from, by default; bind it\n  only when the user asks for it or no acceptable alternative exists. Lanes are\n  **orchestrator-blind**: the same lane fires from every seat the user drives from, and from that\n  CLI's own seat it dispatches the CLI to itself.\n- Surplus placement breaks down when rework and review cost exceed the savings; when the\n  implementer is flaky; when correctness rides on security, concurrency, migrations, or unstated\n  domain knowledge; and when the output **is** the product (debate, architecture, research) —\n  review limits damage, it does not manufacture a good first attempt. Bind those lanes to stronger\n  implementers.\n- An explicit \"spare X\" answer removes X from proposed lanes by default, and overrides blanket\n  posture answers on any lane the user explicitly retains for X — ask whether the posture applies\n  there; omit the dial if unanswered. Never silently stretch one answer across an axis it\n  conflicts with.\n\nQuestion phrasings for the burn\u002Fspare and trust interview live in\n[references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n\nThen propose the lanes. Name them after the work the user described; fall back to `feature`, `tests`,\n`ui`, `fast`, `complex`. Installed implementers only.\n\nShow:\n\n| Lane | Implementer | Model | Effort \u002F variant | Basis | Source (if updating) |\n| --- | --- | --- | --- | --- | --- |\n| feature | opencode | opencode\u002Fgrok | variant: high | your answer + schema requirement | — |\n| tests | codex | — | — | usage data | — |\n| ui | claude | — | — | my opinion (implementer) | — |\n\n**Basis** is mandatory on every lane: `your answer` \u002F `usage data` \u002F `repo` \u002F `my opinion` \u002F\n`schema requirement` (a dial the schema forces is neither evidence nor opinion — say so). A lane you\npicked from model-quality priors is `my opinion` — never present it as something the tooling\ndetermined, and “installed and authenticated” is capability, not evidence of fit. When a lane’s\nimplementer and its dials come from different places, split the label — see\n[references\u002Fsetup-dialogue.md](references\u002Fsetup-dialogue.md).\n\nThen the **complete** JSON (`version`: `delegate-fleet.v1`). One line of why per lane; flag auth or\nmodel uncertainty.\n\nSchema and dial table: [references\u002Fschema.md](references\u002Fschema.md).\n\n### 4. Scope\n\n- User said global \u002F all projects \u002F outside the project → `global`.\n- No git repo → `global` (say so).\n- Else ask once: global vs this repo only.\n\n### 5. Approve and write\n\nOn explicit yes, write **only** the chosen scope (validate first). Build the payload from that\nscope’s raw file (or an empty `lanes` object if new) — not from the effective merged `load` view,\nor a project write will shadow global-only lanes and a global write will promote project-only ones.\n\nCreate a uniquely named file under the platform temporary directory (`$TMPDIR`, `%TEMP%`, or Node\n`os.tmpdir()`; never hard-code `\u002Ftmp`, which breaks on native Windows), write the **exact approved\nJSON** into it with the orchestrator's file-writing tool, and use that populated path as\n`\u003Clanes-json>` below. Never validate an empty temp file. Remove the temp file after the\nvalidation\u002Fwrite attempt, whether it succeeds or fails.\n\n```bash\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" validate \"\u003Clanes-json>\"\nnode \"\u003Cskill-dir>\u002Fscripts\u002Fconfig.mjs\" write --scope global \"\u003Clanes-json>\"\n# or:  write --scope project --cwd \u002Fpath\u002Fto\u002Frepo \"\u003Clanes-json>\"\n```\n\nRe-read with `load`, then confirm the path written and the active lane names. Project writes bind\napproval to the exact config content; later changes fail closed until re-approved. On update, a short\nbefore\u002Fafter is enough.\n\n### 6. Ready to delegate\n\nStop after confirming. Tell the user the map is ready. For later work: read the lane’s\n`implementer`, load that `*-delegate` skill, and dispatch with `--lane \u003Cname>` (explicit\n`--model` \u002F `--effort` \u002F `--variant` still win when passed). Do not start a delegate task\nunless they ask.\n\n## Reconfigure\n\nSame flow. Show the effective current map, propose changes, approve, write one scope’s file.\nReinstalling the skills package must not rewrite these files — they live outside the package.\n\n\n## Limitations\n\n- Never dispatches work itself — only discovers CLIs and writes config after explicit approval.\n- Docs-only import — executable helpers (`scripts\u002F`) not included; see upstream for full runtime. Requires Node 18+.\n\n> Adapted from [amElnagdy\u002Fdelegate-skills](https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills) (MIT) — docs-only, runtime not bundled.\n\n## Configuration trust limits\n\nThe schema reference describes upstream behavior, not an implementation supplied here. Installed\u002Fauthenticated does not establish supported model IDs, quota, safe defaults or enforced read-only behavior. A missing runtime means no project trust hash can be claimed. Before an approved write, inspect symlink destinations, preserve unrelated lanes\u002Fsettings, use private temporary files and avoid exposing session filenames\u002Faccount metadata. A usage scan remains opt-in and requires verifying its actual implementation reads only the promised metadata. No discovery, session scan, config write or delegation occurred during this review.\n",{"name":5,"risk":22,"source":46,"license":14,"category":24,"metadata":47,"date_added":49,"description":7,"source_repo":50,"source_type":27,"compatibility":51,"license_source":52},"https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills",{"version":48},"0.5.0","2026-08-26","amElnagdy\u002Fdelegate-skills","Requires Node 18+. No implementer CLIs are required — the skill discovers what is available.","https:\u002F\u002Fgithub.com\u002FamElnagdy\u002Fdelegate-skills\u002Fblob\u002Fmaster\u002FLICENSE","main","7bf2279c8a65758319526a721d5c4b75683aec9c","skills\u002Fantigravity-delegate-setup",3,"https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections","https:\u002F\u002Fgithub.com\u002Fboisenoise","2026-11-12 14:21:24",{"source":8,"skillName":5,"commands":61,"prompt":71,"zipAvailable":72},[62,65,68],{"tool":63,"command":64},"npx","npx skills add boisenoise\u002Fskills-collections --skill delegate-setup",{"tool":66,"command":67},"bunx","bunx skills add boisenoise\u002Fskills-collections --skill delegate-setup",{"tool":69,"command":70},"pnpm","pnpm dlx skills add boisenoise\u002Fskills-collections --skill delegate-setup","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-delegate-setup\u002FSKILL.md",true,[74,80],{"path":75,"resourceType":76,"resourceTypeDesc":77,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Fschema.md","REFERENCE","参考文档",5131,"https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-delegate-setup\u002Freferences\u002Fschema.md",{"path":81,"resourceType":76,"resourceTypeDesc":77,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Fsetup-dialogue.md",6958,"https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-delegate-setup\u002Freferences\u002Fsetup-dialogue.md",[85,113,138,163,182,205],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":90,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":41},"01M6WNH92BHVFXHGPDZCPJ0AG0","web-access","Use Treg for public web or platform data only when existing search, browser and connected tools cannot cover the task, with explicit cost and data boundaries.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-web-access\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-web-access\u002FSKILL.md",[],[94,95,96,97,98,99,100,101,102,103,104,29,105,106,107,108,109,110,111,112,21],"web","access","web access","official","development","browser","search","data","Tools","read","fetch","youtube","free tools","platform","critical","free","cover","normal","when",{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":117,"sourceHtmlUrl":118,"categories":119,"keywords":125,"publishedAt":41},"01M6WNF9XAYAB2CZDDJBAKNM93","treg","Use the Treg CLI or MCP catalog to discover, price and call authorized external tools with team-scoped credentials and a user-approved budget.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-treg\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-treg\u002FSKILL.md",[120],{"categoryNo":121,"categoryName":122,"taxonomyType":123,"taxonomyTypeDesc":124},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[115,126,127,40,97,98,36,102,128,129,130,131,132,133,134,108,135,136,137,23,21],"MCP","cli","Budget","credentials","team","catalog","use","call","discover","external tools","with","scoped",{"skillNo":139,"name":140,"nameZh":6,"description":141,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":142,"sourceHtmlUrl":143,"categories":144,"keywords":150,"publishedAt":41},"01M6WN93MEAZ4E41DYAHJRPHQT","recruit","Turn a job description into an evidence-linked candidate shortlist through Treg, with job-related ranking and optional authorized business contact enrichment.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-recruit\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-recruit\u002FSKILL.md",[145],{"categoryNo":146,"categoryName":147,"taxonomyType":148,"taxonomyTypeDesc":149},"01JBSEEDCAT000000000000038","人力资源","PROFESSION","适用职业",[140,40,97,36,35,151,152,153,154,155,108,156,157,158,136,159,160,161,162,23,115,21],"email","enrichment","business","evidence","phone","ranking","post","job description","for","person","candidate","reason",{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":167,"sourceHtmlUrl":168,"categories":169,"keywords":170,"publishedAt":41},"01M6WN8R5RXSMHM9B4D9FAFX42","rea-growth","Analyze a company’s public growth and content through Treg, distinguish observations from hypotheses, and produce an evidence-linked playbook within an approved budget.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-rea-growth\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-rea-growth\u002FSKILL.md",[],[171,172,173,40,97,36,174,175,35,176,128,177,178,179,154,108,180,181,23,115,21],"growth","rea","rea growth","produce","content","playbook","marketing","analyze","public","observations","company",{"skillNo":183,"name":184,"nameZh":6,"description":185,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":186,"sourceHtmlUrl":187,"categories":188,"keywords":189,"publishedAt":41},"01M6WN2EVE3RQRGWAKBT4S7YT2","make-ugc","Create original AI UGC-style product videos through Treg: research hooks, approve a licensed presenter and voice, generate clips, add captions and report costs.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-make-ugc\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-make-ugc\u002FSKILL.md",[],[190,191,40,192,97,193,36,194,195,196,197,198,108,199,200,201,202,203,204,23],"make","make ugc","research","create","Video","Captions","ads","short","hooks","head","style","niche","trending","talking-head","approve",{"skillNo":206,"name":207,"nameZh":6,"description":208,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":89,"licenseFileUrl":209,"sourceHtmlUrl":210,"categories":211,"keywords":212,"publishedAt":41},"01M6WN10ARAA63Y8K1YNGXJ2RM","lead-signals","Find and qualify public buying signals through Treg, rank fitting business accounts and optionally enrich authorized business contact information.","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-lead-signals\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fboisenoise\u002Fskills-collections\u002Fblob\u002Fmain\u002Fskills\u002Fantigravity-lead-signals\u002FSKILL.md",[],[213,214,215,40,97,36,216,217,153,218,219,108,220,221,222,223,224,225,226,23],"signals","lead","lead signals","how","find","ready","as","how to","qualify","people","information","enrich","changes","funding"]