[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77MNK4T79JWMTPHA44Y9AYA":3,"skill-same:01M77MNK4T79JWMTPHA44Y9AYA":64},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":32,"ownerDescription":33,"updatedAt":34,"install":35,"resources":47},"01M77MNK4T79JWMTPHA44Y9AYA","setup",null,"Verify the session-flow observer's runtime prerequisites and configuration for this machine, and write the repository's docs\u002Fconventions\u002Fsession-flow.yaml after confirmation. Use when: 'set up session-flow', 'configure the observer', 'is the observer working', 'set worker_continuation for this repo', the SessionStart observer isn't arming, or the observer hook reported a missing prerequisite. Actions: check (read-only, default), apply (writes docs\u002Fconventions\u002Fsession-flow.yaml only). Installs nothing. Re-runnable.","melodic-software\u002Fclaude-code-plugins","melodic-software","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F58273638?v=4","01M77JVQTXNGKCE4MDFY2BC936","Melodic Software","01M77JVQVN9QDWKAH6W03ANP76",30,"MIT","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fskills\u002Fsetup\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\ndescription: \"Verify the session-flow observer's runtime prerequisites and configuration for this machine, and write the repository's docs\u002Fconventions\u002Fsession-flow.yaml after confirmation. Use when: 'set up session-flow', 'configure the observer', 'is the observer working', 'set worker_continuation for this repo', the SessionStart observer isn't arming, or the observer hook reported a missing prerequisite. Actions: check (read-only, default), apply (writes docs\u002Fconventions\u002Fsession-flow.yaml only). Installs nothing. Re-runnable.\"\nargument-hint: \"[check|apply] [\u003Ckey>=\u003Cvalue> ...]\"\nuser-invocable: true\ndisable-model-invocation: true\n---\n\n## Purpose\n\nTwo kinds of configuration. The **detached observer** (see\n[`${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md`](${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md)) has\nruntime prerequisites (Node.js, Python 3.10+, `jq`) and native `userConfig` tunables; setup reports\nboth and writes neither (writing `pluginConfigs` is what the setup contract forbids). The plugin\nalso owns one tracked consumer file, `docs\u002Fconventions\u002Fsession-flow.yaml`, the repository layer of\nthe keys skills read in their own text (keys, values and layers:\n[`${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md`](${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md); schema:\n`${CLAUDE_PLUGIN_ROOT}\u002Fschemas\u002Fsession-flow.schema.json`). `check` reports that file; `apply`\nwrites it and nothing else.\n\nAction routing: no argument or `check` runs the check; `apply` runs the check, then writes.\nRe-running either reads the current state again.\n\n## `check` (read-only)\n\nThe hook and launcher are the single source of truth for what they require and how they degrade:\n`${CLAUDE_PLUGIN_ROOT}\u002Fhooks\u002Fobserver-arm.sh`, `${CLAUDE_PLUGIN_ROOT}\u002Fskills\u002Frunning-retro\u002Fscripts\u002Farm_observer.py`,\nand `observer.py` beside it.\n\n**Read it first.** Probe what it actually does, don't recite this file. Then run each probe via\nBash and report a PASS\u002FFAIL\u002FINFO table with one remediation line per FAIL. Do not modify anything.\n\nWhen the observer is disabled (`observer_enabled` off AND no `arm` invocation in use), every\nprerequisite absence downgrades from FAIL to INFO, the hook exits through its opt-in gate before\ntouching anything, so a deliberately disabled plugin is not broken. Report the probes informationally\nand note that re-enabling restores the FAIL semantics.\n\n1. **Node.js on PATH**, every hook row runs through `node ${CLAUDE_PLUGIN_ROOT}\u002Fhooks\u002Fexec-bash.mjs`.\n   Probe `node --version` through Bash. FAIL if absent: the hook does not launch, so the observer\n   never arms. Remediation: install Node.js on PATH (\u003Chttps:\u002F\u002Fnodejs.org\u002Fen\u002Fdownload>).\n2. **Python 3.10+**, the launcher and tailer are stdlib-only Python 3.10+. Probe the same interpreter\n   detection the hook uses (`python3` then `python`, requiring `sys.version_info >= (3, 10)`). FAIL if\n   none qualifies: without it the hook silently skips arming. Remediation: install Python 3.10+ on PATH.\n3. **`jq`**, the SessionStart hook parses its stdin (`session_id`, `transcript_path`, `source`,\n   `agent_type`) with `jq`. FAIL (auto-arm path only) if absent: the hook exits early and never arms.\n   The manual `arm` action resolves inputs without `jq`, so absence is INFO for that path. Remediation:\n   install `jq` (\u003Chttps:\u002F\u002Fjqlang.org\u002Fdownload\u002F>).\n4. **`claude` CLI on PATH**, the autonomous analysis leg invokes `claude -p`. INFO if absent: the\n   observer still distills and retains observations under its plugin work dir; the analysis run is\n   skipped and nothing is written to the ledger.\n5. **Observer config**. Report the effective value of each native key (an unexpanded `${user_config.…}`\n   token or empty means the default): `${user_config.observer_enabled}` (default off),\n   `${user_config.observer_analysis_enabled}` (default on), `${user_config.observer_analysis_model}`\n   and `${user_config.observer_analysis_effort}` (defaults: read each key's `default` from\n   `${CLAUDE_PLUGIN_ROOT}\u002F.claude-plugin\u002Fplugin.json` `userConfig`, never from memory),\n   `${user_config.observer_analysis_bare}` (default off),\n   `${user_config.observer_idle_seconds}` (default 900), `${user_config.observer_poll_seconds}`\n   (default 5), `${user_config.observer_max_seconds}` (default\n   86400). Call out two hazards: `observer_analysis_bare` on is a FAIL on an OAuth-login install (what bare\n   mode does with OAuth credentials:\n   \u003Chttps:\u002F\u002Fcode.claude.com\u002Fdocs\u002Fen\u002Fheadless#start-faster-with-bare-mode>);\n   `observer_idle_seconds` below the machine's longest expected\n   single turn risks firing analysis on a partial transcript.\n6. **Hook registration**. INFO: confirm the plugin is enabled for this project (`\u002Fplugin` → Installed)\n   rather than parsing settings files. The SessionStart hook only auto-arms when `observer_enabled` is on.\n7. **Repository settings.** Run `node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" --check` from\n   the project root. INFO when `docs\u002Fconventions\u002Fsession-flow.yaml` is absent (every key comes from\n   `userConfig` or its default); PASS with each key's value when the file validates; WARN, never\n   FAIL, when it does not (an invalid value, an unknown or repeated key, an empty or non-scalar\n   value, or a path that is not a plain file), quoting the line the script prints. An invalid\n   value never stops a session-flow skill: the skill names it and drops that layer. An unknown key\n   is named and ignored, and the valid keys still print. `apply` overwrites only a value that is\n   out of the list, empty or null; it refuses every other shape (a repeated key, an unknown key\n   holding a map or list, a map or list, an empty quoted string), which the operator fixes by\n   hand.\n\n## `apply` (writes `docs\u002Fconventions\u002Fsession-flow.yaml` only)\n\n1. Run `check` and show its table.\n2. **Resolve the values.** With complete `\u003Ckey>=\u003Cvalue>` arguments, use them. Otherwise ask one key\n   at a time, recommendation first, from the table in `${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md`\n   (`worker_continuation`: `resume`, the default, unless the team wants each new unit in a fresh\n   worker; `wip_commit`: `false`, the default, unless the team wants a pause\n   handoff to make one local `chore(wip):` commit, an unquoted `true` or `false`;\n   `transcript_scope`: leave it unset unless the team wants to narrow every member's\n   `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall` scans to `worktree` or `repo`, since\n   this key can only narrow a user's option, never widen it). Never invent a key the schema does not list.\n3. **Write.** One call with every value:\n\n   ```bash\n   node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" worker_continuation=respawn\n   ```\n\n   Pass a key the operator accepts at the plugin default as `--default \u003Ckey>` instead of\n   `\u003Ckey>=\u003Cvalue>`, so a later change to that default reaches it; `\u003Ckey>=\u003Cvalue>` records the\n   team's choice, which the script never changes. The script also records `written_against` and\n   `provenance:`, so a file with no record, or one written against another plugin version, shows a\n   diff even when no value changes.\n\n   The script checks each value against the schema, refuses a key given twice, validates the\n   existing file first (refusing it unless every problem is a value out of the list, empty or\n   null, or an unknown key it names in one warning and keeps as written), and validates the whole\n   resulting file (no repeated key, one non-empty scalar per key) before it writes; any refusal\n   exits 1 and writes nothing. It writes only `\u003Cgit toplevel>\u002Fdocs\u002Fconventions\u002Fsession-flow.yaml`:\n   it resolves a symlinked root first, then refuses a root that is `$HOME` or an ancestor of it, a\n   symlink on that path, on `docs` or on `docs\u002Fconventions`, a directory resolving outside the\n   repository, and a target with more than one hard link, and writes through a temp file renamed\n   into place. A missing file is created. A value already in place prints `already configured` and\n   writes nothing.\n4. **An existing file that would change** exits 3 and prints a unified diff without writing. Show\n   the operator that diff and ask whether to write it. Only on an explicit yes, re-run the same\n   call with `--yes`; on anything else, stop with the file unchanged. A request to \"set it\" is not\n   a yes to a diff the operator has not seen.\n   When the convention-home offer prints below the diff, ask about it too (the next step) before\n   any re-run, since `--yes` writes both.\n5. **The convention-home offer.** When the repository's `AGENTS.md` or `CLAUDE.md` binds no\n   convention home, the script prints an offer: with the operator's confirmation it adds a\n   convention-home pointer region to that instruction file (creating `AGENTS.md` when neither\n   exists or `CLAUDE.md` only imports it) and the personal override ignore line\n   `docs\u002Fconventions\u002F**\u002F*.local.*` to `.gitignore`, the only files `apply` writes outside\n   `docs\u002Fconventions\u002F`. It prints the offer at the end of an exit 0 run and below the diff of an\n   exit 3 run, so show the operator stdout on exit 0 too. Ask whether to add the region and the\n   ignore line. Only on an explicit yes, re-run the same call with `--yes`, which also writes any\n   diff shown with it. On a no, run `node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" --decline`\n   alone: it records the decline in the instruction file, and the next run does not offer again.\n   To write a shown diff without the region, run `--decline` first, then the same call with\n   `--yes`. When the instruction file or `.gitignore` is a symlink or hard link the script will\n   not write through, it prints a one-line refusal in place of the offer, with the same exit\n   code: show it, and do not re-run with `--yes`, which that refusal stops with exit 1. A refused\n   instruction file refuses `--decline` too, so the operator fixes the link or adds the region by\n   hand.\n6. **Verify.** Re-run `check` and report the value from its table, not from the write. Then the\n   tracked-file pair: `git check-ignore -v docs\u002Fconventions\u002Fsession-flow.yaml` reports no match (a\n   match means the team never receives the file: say so, and leave `.gitignore` to the operator),\n   and `git ls-files --error-unmatch docs\u002Fconventions\u002Fsession-flow.yaml` exits 0. Non-zero right\n   after a fresh write means \"written but untracked: commit it to share with the team\", never\n   success.\n\n## Remediation guidance (printed by `check`; the operator applies it)\n\nFor the observer there is no write path. For each FAIL, `check` closes by offering the remediation: install the missing\ntool, or route observer reconfiguration through Claude Code's native flow.\nDo not write the plugin cache, Claude Code user settings, or `pluginConfigs`.\n\nReconfigure the observer's `userConfig` keys through Claude Code's native flow, interactively with\n`\u002Fplugin configure session-flow@\u003Cmarketplace>` or by a headless route, as the marketplace's\nplugin-reconfiguration convention gives them in\n[§ The routes](..\u002F..\u002Freference\u002Fplugin-reconfiguration.md#the-routes)\n(it owns the verified-version record). Apply the scope and failure-signal caveats in its\n[§ Caveats every setup skill's short form carries](..\u002F..\u002Freference\u002Fplugin-reconfiguration.md#caveats-every-setup-skills-short-form-carries).\nRead both sections and print what they give, the fallback route, its behavior on an\nalready-installed plugin and the scope to pass included, rather than only naming them, and\nleave out the verified-version record.\nNever uninstall to reconfigure; that section links what an uninstall deletes. Afterwards rerun\n`check` in a **fresh session**, since a same-session `check` can still show the old value, and\nreport the observed effective value, never an unobserved change.\n\n- **Pointer**: for the headless command's flags, how it reports a saved or refused value, and\n  when a saved value applies, fetch\n  [plugin CLI reference § plugin configure](https:\u002F\u002Fcode.claude.com\u002Fdocs\u002Fen\u002Fplugins\u002Fcli-reference#plugin-configure)\n  live.\n\n## Output\n\n`check`: the PASS\u002FFAIL\u002FINFO\u002FWARN table with one remediation line per FAIL. `apply`: the table\nbefore and after, the diff when one was shown, the written path, and whether the file is tracked.\n\n## Next\n\n`\u002Fsession-flow:orchestrate`, which reads `worker_continuation`; `\u002Fsession-flow:retro`, whose session mode\nreads `retro_lenses`; `\u002Fsession-flow:handoff`, which reads\n`wip_commit`; `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall`, which read `transcript_scope`;\n`\u002Fsession-flow:recall`, which also reads `recall_report_kinds` and `recall_doc_paths`.\n\n## Gotchas\n\n- **Prerequisites are the observer's, plus `jq` for retro's docs reads and `gh` for\n  `\u002Fsession-flow:recall`'s pull request reads.**\n  `\u002Fsession-flow:find-handoff` uses Python 3.10+ when present and lists fewer transcript\n  directories without it. The other session-flow skills need no installed tool. The repository settings are `worker_continuation`,\n  read by `\u002Fsession-flow:orchestrate`, `retro_lenses`, read by `\u002Fsession-flow:retro` session mode,\n  `wip_commit`, read by `\u002Fsession-flow:handoff`, and\n  `transcript_scope`, read by `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall`, where the\n  narrower layer wins, and `recall_report_kinds` and `recall_doc_paths`, read by\n  `\u002Fsession-flow:recall`.\n- **The repository file reaches the team only once committed.** `apply` leaves it uncommitted on\n  purpose, and the tracked-file pair says so.\n- **`observer_analysis_bare` and auth.** Leave it off unless auth is an env-var API key (why: the\n  bare-mode pointer in `check` step 5). Full detail in\n  `${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md`.\n","## Purpose\n\nTwo kinds of configuration. The **detached observer** (see\n[`${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md`](${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md)) has\nruntime prerequisites (Node.js, Python 3.10+, `jq`) and native `userConfig` tunables; setup reports\nboth and writes neither (writing `pluginConfigs` is what the setup contract forbids). The plugin\nalso owns one tracked consumer file, `docs\u002Fconventions\u002Fsession-flow.yaml`, the repository layer of\nthe keys skills read in their own text (keys, values and layers:\n[`${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md`](${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md); schema:\n`${CLAUDE_PLUGIN_ROOT}\u002Fschemas\u002Fsession-flow.schema.json`). `check` reports that file; `apply`\nwrites it and nothing else.\n\nAction routing: no argument or `check` runs the check; `apply` runs the check, then writes.\nRe-running either reads the current state again.\n\n## `check` (read-only)\n\nThe hook and launcher are the single source of truth for what they require and how they degrade:\n`${CLAUDE_PLUGIN_ROOT}\u002Fhooks\u002Fobserver-arm.sh`, `${CLAUDE_PLUGIN_ROOT}\u002Fskills\u002Frunning-retro\u002Fscripts\u002Farm_observer.py`,\nand `observer.py` beside it.\n\n**Read it first.** Probe what it actually does, don't recite this file. Then run each probe via\nBash and report a PASS\u002FFAIL\u002FINFO table with one remediation line per FAIL. Do not modify anything.\n\nWhen the observer is disabled (`observer_enabled` off AND no `arm` invocation in use), every\nprerequisite absence downgrades from FAIL to INFO, the hook exits through its opt-in gate before\ntouching anything, so a deliberately disabled plugin is not broken. Report the probes informationally\nand note that re-enabling restores the FAIL semantics.\n\n1. **Node.js on PATH**, every hook row runs through `node ${CLAUDE_PLUGIN_ROOT}\u002Fhooks\u002Fexec-bash.mjs`.\n   Probe `node --version` through Bash. FAIL if absent: the hook does not launch, so the observer\n   never arms. Remediation: install Node.js on PATH (\u003Chttps:\u002F\u002Fnodejs.org\u002Fen\u002Fdownload>).\n2. **Python 3.10+**, the launcher and tailer are stdlib-only Python 3.10+. Probe the same interpreter\n   detection the hook uses (`python3` then `python`, requiring `sys.version_info >= (3, 10)`). FAIL if\n   none qualifies: without it the hook silently skips arming. Remediation: install Python 3.10+ on PATH.\n3. **`jq`**, the SessionStart hook parses its stdin (`session_id`, `transcript_path`, `source`,\n   `agent_type`) with `jq`. FAIL (auto-arm path only) if absent: the hook exits early and never arms.\n   The manual `arm` action resolves inputs without `jq`, so absence is INFO for that path. Remediation:\n   install `jq` (\u003Chttps:\u002F\u002Fjqlang.org\u002Fdownload\u002F>).\n4. **`claude` CLI on PATH**, the autonomous analysis leg invokes `claude -p`. INFO if absent: the\n   observer still distills and retains observations under its plugin work dir; the analysis run is\n   skipped and nothing is written to the ledger.\n5. **Observer config**. Report the effective value of each native key (an unexpanded `${user_config.…}`\n   token or empty means the default): `${user_config.observer_enabled}` (default off),\n   `${user_config.observer_analysis_enabled}` (default on), `${user_config.observer_analysis_model}`\n   and `${user_config.observer_analysis_effort}` (defaults: read each key's `default` from\n   `${CLAUDE_PLUGIN_ROOT}\u002F.claude-plugin\u002Fplugin.json` `userConfig`, never from memory),\n   `${user_config.observer_analysis_bare}` (default off),\n   `${user_config.observer_idle_seconds}` (default 900), `${user_config.observer_poll_seconds}`\n   (default 5), `${user_config.observer_max_seconds}` (default\n   86400). Call out two hazards: `observer_analysis_bare` on is a FAIL on an OAuth-login install (what bare\n   mode does with OAuth credentials:\n   \u003Chttps:\u002F\u002Fcode.claude.com\u002Fdocs\u002Fen\u002Fheadless#start-faster-with-bare-mode>);\n   `observer_idle_seconds` below the machine's longest expected\n   single turn risks firing analysis on a partial transcript.\n6. **Hook registration**. INFO: confirm the plugin is enabled for this project (`\u002Fplugin` → Installed)\n   rather than parsing settings files. The SessionStart hook only auto-arms when `observer_enabled` is on.\n7. **Repository settings.** Run `node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" --check` from\n   the project root. INFO when `docs\u002Fconventions\u002Fsession-flow.yaml` is absent (every key comes from\n   `userConfig` or its default); PASS with each key's value when the file validates; WARN, never\n   FAIL, when it does not (an invalid value, an unknown or repeated key, an empty or non-scalar\n   value, or a path that is not a plain file), quoting the line the script prints. An invalid\n   value never stops a session-flow skill: the skill names it and drops that layer. An unknown key\n   is named and ignored, and the valid keys still print. `apply` overwrites only a value that is\n   out of the list, empty or null; it refuses every other shape (a repeated key, an unknown key\n   holding a map or list, a map or list, an empty quoted string), which the operator fixes by\n   hand.\n\n## `apply` (writes `docs\u002Fconventions\u002Fsession-flow.yaml` only)\n\n1. Run `check` and show its table.\n2. **Resolve the values.** With complete `\u003Ckey>=\u003Cvalue>` arguments, use them. Otherwise ask one key\n   at a time, recommendation first, from the table in `${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fconfig.md`\n   (`worker_continuation`: `resume`, the default, unless the team wants each new unit in a fresh\n   worker; `wip_commit`: `false`, the default, unless the team wants a pause\n   handoff to make one local `chore(wip):` commit, an unquoted `true` or `false`;\n   `transcript_scope`: leave it unset unless the team wants to narrow every member's\n   `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall` scans to `worktree` or `repo`, since\n   this key can only narrow a user's option, never widen it). Never invent a key the schema does not list.\n3. **Write.** One call with every value:\n\n   ```bash\n   node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" worker_continuation=respawn\n   ```\n\n   Pass a key the operator accepts at the plugin default as `--default \u003Ckey>` instead of\n   `\u003Ckey>=\u003Cvalue>`, so a later change to that default reaches it; `\u003Ckey>=\u003Cvalue>` records the\n   team's choice, which the script never changes. The script also records `written_against` and\n   `provenance:`, so a file with no record, or one written against another plugin version, shows a\n   diff even when no value changes.\n\n   The script checks each value against the schema, refuses a key given twice, validates the\n   existing file first (refusing it unless every problem is a value out of the list, empty or\n   null, or an unknown key it names in one warning and keeps as written), and validates the whole\n   resulting file (no repeated key, one non-empty scalar per key) before it writes; any refusal\n   exits 1 and writes nothing. It writes only `\u003Cgit toplevel>\u002Fdocs\u002Fconventions\u002Fsession-flow.yaml`:\n   it resolves a symlinked root first, then refuses a root that is `$HOME` or an ancestor of it, a\n   symlink on that path, on `docs` or on `docs\u002Fconventions`, a directory resolving outside the\n   repository, and a target with more than one hard link, and writes through a temp file renamed\n   into place. A missing file is created. A value already in place prints `already configured` and\n   writes nothing.\n4. **An existing file that would change** exits 3 and prints a unified diff without writing. Show\n   the operator that diff and ask whether to write it. Only on an explicit yes, re-run the same\n   call with `--yes`; on anything else, stop with the file unchanged. A request to \"set it\" is not\n   a yes to a diff the operator has not seen.\n   When the convention-home offer prints below the diff, ask about it too (the next step) before\n   any re-run, since `--yes` writes both.\n5. **The convention-home offer.** When the repository's `AGENTS.md` or `CLAUDE.md` binds no\n   convention home, the script prints an offer: with the operator's confirmation it adds a\n   convention-home pointer region to that instruction file (creating `AGENTS.md` when neither\n   exists or `CLAUDE.md` only imports it) and the personal override ignore line\n   `docs\u002Fconventions\u002F**\u002F*.local.*` to `.gitignore`, the only files `apply` writes outside\n   `docs\u002Fconventions\u002F`. It prints the offer at the end of an exit 0 run and below the diff of an\n   exit 3 run, so show the operator stdout on exit 0 too. Ask whether to add the region and the\n   ignore line. Only on an explicit yes, re-run the same call with `--yes`, which also writes any\n   diff shown with it. On a no, run `node \"${CLAUDE_SKILL_DIR}\u002Fscripts\u002Fsetup-apply.mjs\" --decline`\n   alone: it records the decline in the instruction file, and the next run does not offer again.\n   To write a shown diff without the region, run `--decline` first, then the same call with\n   `--yes`. When the instruction file or `.gitignore` is a symlink or hard link the script will\n   not write through, it prints a one-line refusal in place of the offer, with the same exit\n   code: show it, and do not re-run with `--yes`, which that refusal stops with exit 1. A refused\n   instruction file refuses `--decline` too, so the operator fixes the link or adds the region by\n   hand.\n6. **Verify.** Re-run `check` and report the value from its table, not from the write. Then the\n   tracked-file pair: `git check-ignore -v docs\u002Fconventions\u002Fsession-flow.yaml` reports no match (a\n   match means the team never receives the file: say so, and leave `.gitignore` to the operator),\n   and `git ls-files --error-unmatch docs\u002Fconventions\u002Fsession-flow.yaml` exits 0. Non-zero right\n   after a fresh write means \"written but untracked: commit it to share with the team\", never\n   success.\n\n## Remediation guidance (printed by `check`; the operator applies it)\n\nFor the observer there is no write path. For each FAIL, `check` closes by offering the remediation: install the missing\ntool, or route observer reconfiguration through Claude Code's native flow.\nDo not write the plugin cache, Claude Code user settings, or `pluginConfigs`.\n\nReconfigure the observer's `userConfig` keys through Claude Code's native flow, interactively with\n`\u002Fplugin configure session-flow@\u003Cmarketplace>` or by a headless route, as the marketplace's\nplugin-reconfiguration convention gives them in\n[§ The routes](..\u002F..\u002Freference\u002Fplugin-reconfiguration.md#the-routes)\n(it owns the verified-version record). Apply the scope and failure-signal caveats in its\n[§ Caveats every setup skill's short form carries](..\u002F..\u002Freference\u002Fplugin-reconfiguration.md#caveats-every-setup-skills-short-form-carries).\nRead both sections and print what they give, the fallback route, its behavior on an\nalready-installed plugin and the scope to pass included, rather than only naming them, and\nleave out the verified-version record.\nNever uninstall to reconfigure; that section links what an uninstall deletes. Afterwards rerun\n`check` in a **fresh session**, since a same-session `check` can still show the old value, and\nreport the observed effective value, never an unobserved change.\n\n- **Pointer**: for the headless command's flags, how it reports a saved or refused value, and\n  when a saved value applies, fetch\n  [plugin CLI reference § plugin configure](https:\u002F\u002Fcode.claude.com\u002Fdocs\u002Fen\u002Fplugins\u002Fcli-reference#plugin-configure)\n  live.\n\n## Output\n\n`check`: the PASS\u002FFAIL\u002FINFO\u002FWARN table with one remediation line per FAIL. `apply`: the table\nbefore and after, the diff when one was shown, the written path, and whether the file is tracked.\n\n## Next\n\n`\u002Fsession-flow:orchestrate`, which reads `worker_continuation`; `\u002Fsession-flow:retro`, whose session mode\nreads `retro_lenses`; `\u002Fsession-flow:handoff`, which reads\n`wip_commit`; `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall`, which read `transcript_scope`;\n`\u002Fsession-flow:recall`, which also reads `recall_report_kinds` and `recall_doc_paths`.\n\n## Gotchas\n\n- **Prerequisites are the observer's, plus `jq` for retro's docs reads and `gh` for\n  `\u002Fsession-flow:recall`'s pull request reads.**\n  `\u002Fsession-flow:find-handoff` uses Python 3.10+ when present and lists fewer transcript\n  directories without it. The other session-flow skills need no installed tool. The repository settings are `worker_continuation`,\n  read by `\u002Fsession-flow:orchestrate`, `retro_lenses`, read by `\u002Fsession-flow:retro` session mode,\n  `wip_commit`, read by `\u002Fsession-flow:handoff`, and\n  `transcript_scope`, read by `\u002Fsession-flow:find-handoff` and `\u002Fsession-flow:recall`, where the\n  narrower layer wins, and `recall_report_kinds` and `recall_doc_paths`, read by\n  `\u002Fsession-flow:recall`.\n- **The repository file reaches the team only once committed.** `apply` leaves it uncommitted on\n  purpose, and the tracked-file pair says so.\n- **`observer_analysis_bare` and auth.** Leave it off unless auth is an env-var API key (why: the\n  bare-mode pointer in `check` step 5). Full detail in\n  `${CLAUDE_PLUGIN_ROOT}\u002Freference\u002Fobserver.md`.\n",{"description":7,"argument-hint":24,"user-invocable":25,"disable-model-invocation":25},"[check|apply] [\u003Ckey>=\u003Cvalue> ...]",true,"main","4f2dcaaff35e8f9fccc6a9f4d82a5830ab17a5c5","plugins\u002Fsession-flow\u002Fskills\u002Fsetup",7,"https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins","https:\u002F\u002Fgithub.com\u002Fmelodic-software","https:\u002F\u002Fmelodicsoftware.com","Independent software studio building developer tools, automation, and application services, with roots in music.","2026-11-15 05:15:06",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":25},[37,40,43],{"tool":38,"command":39},"npx","npx skills add melodic-software\u002Fclaude-code-plugins --skill setup",{"tool":41,"command":42},"bunx","bunx skills add melodic-software\u002Fclaude-code-plugins --skill setup",{"tool":44,"command":45},"pnpm","pnpm dlx skills add melodic-software\u002Fclaude-code-plugins --skill setup","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fskills\u002Fsetup\u002FSKILL.md",[48,54,60],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"evals\u002Fevals.json","REFERENCE","参考文档",6280,"https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fskills\u002Fsetup\u002Fevals\u002Fevals.json",{"path":55,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":58,"sourceHtmlUrl":59},"scripts\u002Fsetup-apply.mjs","SCRIPT","脚本",1147,"https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fskills\u002Fsetup\u002Fscripts\u002Fsetup-apply.mjs",{"path":61,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":62,"sourceHtmlUrl":63},"scripts\u002Fsetup-apply.test.sh",18643,"https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fskills\u002Fsetup\u002Fscripts\u002Fsetup-apply.test.sh",[65,72,79,86,91,98],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7DYYDG31B9YX7N9MB9QW2GW","track","Backlog reads through the bound tracker: search open and closed items, view one item with its comments and state history. Use when another skill searches or reads tracker items.","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fsession-flow\u002Fevals\u002Frecall-still-reports-reopened-bug\u002Fwork-items-stub\u002Fskills\u002Ftrack\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M7DYXVBAVG3NV30PA7GKW9Q6","reviewer-media","Reviewer media for one pull request: decides whether it needs a narrated walkthrough, builds the storyboard from \u002Freview:explain-change's digest, has \u002Fexplainer-video:produce render it and hands the manifest to \u002Fsource-control:pull-request media. Use when asked for 'reviewer media', 'a walkthrough video for this PR', 'a produced video of this change', or when pull-request ready or monitor hands a PR over.","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Freview\u002Fskills\u002Freviewer-media\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7DYRZ5YD44AMYH26YR25GZ8","route","Model and effort per agent role (orchestrator, worker, verifier, retrieval) for a multi-agent run, resolved through the bundled defaults and the user, team and personal override config layers into fenced JSON with single-agent and fan-out variants and the layer behind each value. Use when asked 'which model for the subagents', 'route the roles' or 'model routing', or before a skill passes a role's model on a spawn.","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fdiscovery\u002Fevals\u002Fex-model-pin-without-override\u002Fmulti-agent-stub\u002Fskills\u002Froute\u002FSKILL.md",[],[],{"skillNo":87,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7DYRTVT4FWTGDJY0HCCFJMZ","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fdiscovery\u002Fevals\u002Fex-model-override-from-route\u002Fmulti-agent-stub\u002Fskills\u002Froute\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7DYNZC8PNW3T27V2ADKECKJ","principles","Module-design vocabulary for architecture findings, and the checks built on it. Use when: 'what is a deep module', 'what is a seam', 'deep vs shallow module', 'what is the deletion test', 'leverage and locality', 'module design vocabulary', 'which word do we use for X in architecture findings'. Scanning a codebase for friction is \u002Farchitecture:improve.","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Farchitecture\u002Fskills\u002Fprinciples\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7DYNMCZNEWBFHV9KTSS6H60","write","Structured five-field bug report (title, steps to reproduce, expected vs actual, justified severity, suggested fix location) from an informal description, read-only. Use when the user names a defect they observed, or asks for one written up ('report a bug', 'file a bug', 'write this up as a bug').","https:\u002F\u002Fgithub.com\u002Fmelodic-software\u002Fclaude-code-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Farchitecture\u002Fevals\u002Fdeepening-defect-goes-to-bugs-write\u002Fbugs-stub\u002Fskills\u002Fwrite\u002FSKILL.md",[],[]]