[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6NQS7ER5H3THFFTJMB6S605":3,"skill-same:01M6NQS7ER5H3THFFTJMB6S605":182},{"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":23,"publishedAt":44,"linkOnly":45,"markdownContent":46,"markdownBody":47,"frontMatter":48,"refName":99,"commitSha":100,"skillDirectory":101,"forksCount":14,"repositoryHtmlUrl":102,"repositoryHomepageUrl":103,"ownerHtmlUrl":104,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":105,"install":106,"resources":119},"01M6NQS7ER5H3THFFTJMB6S605","team-manager",null,"Create and maintain TeamMember entities — humans and named AI personas that participate in the project with persistent identity, personality, and tiered memory. Use when adding a new persistent team-member (human or named AI), managing the name pool, scaffolding memory trees, or exporting\u002Fimporting a team-member bundle. Replaces the deprecated actor-profile skill (DEC-20260422_0233-SpryTulip).","projectious-work\u002Fprocesskit","projectious-work","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F263889726?v=4","01M6N97TJJ118WQ35CHRV8RD0E","projectious.work","01M6N97TPWJPF0BMKYQ4PBY5W8",0,"MIT","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"team manager","manager","team","references","fresh","get","project","scaffolding","query","maintain","ai-agent","memory","default","log","work","subagent","sqlite","SET","create","init","2026-11-10 10:12:30",false,"---\nname: team-manager\ndescription: |\n  Create and maintain TeamMember entities — humans and named AI personas that participate in the project with persistent identity, personality, and tiered memory. Use when adding a new persistent team-member (human or named AI), managing the name pool, scaffolding memory trees, or exporting\u002Fimporting a team-member bundle. Replaces the deprecated actor-profile skill (DEC-20260422_0233-SpryTulip).\nmetadata:\n  processkit:\n    apiVersion: processkit.projectious.work\u002Fv2\n    id: SKILL-team-manager\n    version: \"1.0.0\"\n    created: 2026-04-22T00:00:00Z\n    category: processkit\n    layer: 1\n    replaces:\n      - SKILL-actor-profile\n    uses:\n      - skill: event-log\n        purpose: Log events to keep the audit trail accurate after every write.\n      - skill: index-management\n        purpose: Query existing entities and keep the SQLite index fresh after writes.\n      - skill: id-management\n        purpose: Allocate unique entity identifiers via central ID generation.\n      - skill: role-management\n        purpose: Resolve default_role references against the Role catalog.\n    provides:\n      primitives: [TeamMember]\n      mcp_tools:\n        - create_team_member\n        - get_team_member\n        - list_team_members\n        - get_active_interlocutor\n        - set_active_interlocutor\n        - update_team_member\n        - deactivate_team_member\n        - reactivate_team_member\n        - reserve_name\n        - release_name\n        - list_available_names\n        - suggest_name\n        - init_memory_tree\n        - export_team_member\n        - export_claude_subagent\n        - export_claude_subagents\n        - import_team_member\n        - check_consistency\n        - check_all_consistency\n      templates: [team-member-human, team-member-ai-agent]\n---\n\n# Team Manager\n\n## Intro\n\nTeamMembers are the persistent participants in the project — named AI\npersonas, services, consultants, and explicitly managed human aliases —\nthat accumulate identity, personality, and tiered memory over time. This\nskill owns their lifecycle: creation, memory tree scaffolding, a curated\ninternational name pool, A2A-compatible Agent Cards, export\u002Fimport\nbundles, and consistency checks.\n\nUnlike ephemeral worker invocations (role+seniority dispatches resolved on\nthe fly and never persisted), TeamMembers live as directory trees on disk:\n`persona.md` + `card.json` + `team-member.md` (entity) + tiered memory\nsubdirectories + optional `private\u002F` (gitignored).\n\nHuman TeamMembers are privacy-sensitive. Do not auto-create them from\nambient host, git, or harness identity. `create_team_member(type=\"human\",\n...)` stores only the slug as the committed display name by default and\ndrops `email`\u002F`handle`; pass `allow_committed_pii=true` only after the\nhuman explicitly agrees that their name or contact details may live in\nrepo-visible files. Store local-only human details under `private\u002F`, not\nin `team-member.md`.\n\n> **MCP server.** This skill ships a self-contained MCP server at\n> `mcp\u002Fserver.py` (PEP 723 script — requires `uv` and Python ≥ 3.10 on\n> PATH). Agent harnesses reach its tools by reading a single MCP config\n> file at startup, so the contents of `mcp\u002Fmcp-config.json` must be merged\n> into the harness's MCP config and placed at the harness-specific path\n> before this skill is usable.\n\n## Overview\n\n### Invocation surface\n\nThis skill is not a CLI; it is reached entirely via MCP tools served\nby `mcp\u002Fserver.py`. Tools cluster into six capability groups\n(lifecycle, session identity, name pool, memory scaffolding,\nexport\u002Fimport, consistency). Every write tool logs an `event-log`\nentry and updates the SQLite index automatically — callers do not log\nor reindex by hand. Read tools (`get_team_member`,\n`list_team_members`, `get_active_interlocutor`,\n`get_interlocutor_runtime_binding`, `list_available_names`,\n`suggest_name`, `get_team_member_runtime`, `list_team_member_runtimes`,\n`check_consistency`) are side-effect-free and safe to call from any\nagent context.\n\n### Capability groups\n\n| Group | Tools | Use when |\n|---|---|---|\n| Lifecycle | `create_team_member`, `get_team_member`, `list_team_members`, `update_team_member`, `deactivate_team_member`, `reactivate_team_member` | Adding, editing, retiring or restoring a persistent team-member entity |\n| Session identity | `get_active_interlocutor`, `get_interlocutor_runtime_binding`, `set_active_interlocutor` | Showing which persistent TeamMember, if any, is the current harness interlocutor, plus the resolved runtime binding when available |\n| Runtime launch | `launch_team_member`, `launch_workitem_assignee`, `get_team_member_runtime`, `list_team_member_runtimes`, `stop_team_member_runtime` | Creating and tracking a harness-consumable TeamMember runtime request with explicit model, status, and write policy |\n| Name pool | `reserve_name`, `release_name`, `list_available_names`, `suggest_name` | Choosing or rotating an AI persona name; the pool is the canonical source of truth for AI-agent slugs |\n| Memory scaffolding | `init_memory_tree` | Creating the six tier subdirectories + frontmatter-validated headers under `context\u002Fteam-members\u002F\u003Cslug>\u002F` |\n| Export \u002F import | `export_team_member`, `export_claude_subagent`, `export_claude_subagents`, `import_team_member` | Moving a TeamMember between projects or generating provider-specific adapter files; tarball export redacts confidential\u002FPII memory |\n| Consistency | `check_consistency`, `check_all_consistency` | Pre-release health-check; CI validation; ad-hoc audits, including human committed-PII warnings |\n\n### Common workflows\n\n**Adding a human contributor.** Prefer an ephemeral Actor\u002Freference unless\nthe project really needs a persistent human TeamMember. When persistence\nis needed, use a privacy-preserving slug such as `human-user` or\n`project-owner`: `create_team_member(type=\"human\", name=\"human-user\",\nslug=\"human-user\", ...)` → `init_memory_tree(slug)`. No name-pool\ninteraction; humans use any valid kebab-case slug. Real names, emails,\nand handles require `allow_committed_pii=true` and explicit human consent.\n\n**Adding a named AI persona.** `suggest_name(kind=\"neutral\")` →\n`create_team_member(type=\"ai-agent\", slug=\u003Cname>, ...)` (auto-reserves\nthe name) → `init_memory_tree(slug)` → write `persona.md` and the\nA2A-compliant `card.json`. The name is now drawn from the pool and\nmust remain in `data\u002Fname-pool.yaml` until released.\n\n**Retiring a persona.** `deactivate_team_member(slug, reason)` (sets\n`active: false`; preserves history) → optionally\n`release_name(name)` to return the slug to the pool. To bring the\npersona back, call `reactivate_team_member(slug)`.\n\n**Migrating a persona between projects.**\n`export_team_member(slug)` → ship the tarball → at the destination\n`import_team_member(tarball_path)`. Schema and Agent Card are\nre-validated on import; signature presence is required (crypto\nverification is deferred).\n\n**Showing the active interlocutor.**\n`set_active_interlocutor(id)` writes\n`context\u002Fteam\u002Fsession-identity.json` with the TeamMember that should be\npresented as the harness's main speaker. `get_active_interlocutor()`\nreturns a compact label and speaker prefix for status briefings and\nsession-start UI. This is session identity, not lifecycle state:\n`TeamMember.spec.active` continues to mean whether a TeamMember is\neligible to participate at all.\n\nUse `get_interlocutor_runtime_binding(observed_model?,\nobserved_effort?)` when the caller can supply the harness's actual\nruntime. The tool returns the provider-neutral TeamMember identity, the\nmodel-recommender candidate after local provider gates, the negotiated\nharness mode, and an informational mismatch report. processkit cannot\nhot-swap an already running primary harness session; unsupported\nharnesses therefore fall back to identity-only behavior while still\ngrowing memory on the configured TeamMember.\n\n**Launching TeamMember runtimes.**\n`launch_team_member(id, harness?, ...)` creates a runtime record under\n`context\u002Fteam\u002Fruntime-sessions.json` and returns a harness-specific\ndispatch payload. `launch_workitem_assignee(workitem_id, ...)` resolves\nthe WorkItem's `TEAMMEMBER-*` assignee first and then records the same\nruntime contract. Runtime state is separate from WorkItem state:\n`queued`, `running`, `failed`, and `stopped` describe the harness\nsession, not the work lifecycle.\n\nThe MCP server does not spawn host harness processes directly. For\nClaude, the dispatch payload names the `subagent_type` and refreshes the\n`.claude\u002Fagents\u002F\u003Cslug>.md` adapter by default. For Codex, Aider, and\nOpenCode, the payload carries the TeamMember identity, resolved model,\neffort, write scope, and MCP\u002Fcontext-write policy for the outer harness\nadapter to consume. Use `list_team_member_runtimes(active_only=true)`\nand `stop_team_member_runtime(handle)` to report and update status.\n\n**Exporting Claude Code subagents.**\n`export_claude_subagent(slug)` writes one generated adapter to\n`.claude\u002Fagents\u002F\u003Cslug>.md`; `export_claude_subagents()` exports all\nactive non-human TeamMembers by default. The generated file uses\nClaude Code's Markdown + YAML frontmatter format, sets `model:\ninherit`, and omits `tools` so Claude Code inherits the main session's\nmodel and permissions. Pass `model_policy=\"resolved\"` to emit a Claude\nmodel and effort only when processkit routing resolves to an Anthropic\ncandidate. TeamMember remains the canonical, provider-neutral identity;\n`.claude\u002Fagents\u002F*.md` is only an adapter.\n\n**Pre-release audit.** `check_all_consistency()` → triage findings by\nseverity. Errors block release; warnings are advisory. Treat\n`team.privacy.committed_pii` as release-blocking for derived-project\ntemplates and installer output: rotate the TeamMember to an alias or move\nthe personal data into `private\u002F`.\n\n### Mental model\n\nA TeamMember is **a directory tree on disk**, not a row in a database.\nThe entity file (`team-member.md` \u002F `team-member.yaml` frontmatter)\nis the SQLite-indexed object; `persona.md`, `card.json`, and the six\ntier subdirectories live alongside it as durable, human-editable\ncontext. Humans\u002Fservices have no name-pool requirement; AI personas\ndo. TeamMembers are distinct from **ephemeral worker invocations** —\nthose are resolved on the fly by `task-router` against role+seniority\nand never persist as TeamMembers.\n\n### Pointers to the full reference\n\nThe `## Full reference` section below covers the MCP tool contract,\nthe SQLite-index schema fields, the `data\u002Fname-pool.yaml` document\nformat, the consistency check codes in greater detail, the export\nbundle layout, the migration path from `actor-profile`\n(DEC-20260422_0233-SpryTulip), and the corner cases for memory\nsensitivity tagging. Read it before extending the skill or changing\nany tool's contract.\n\n## ID convention\n\nTeamMember IDs follow a single shape:\n\n    TEAMMEMBER-\u003Cslug>\n\nwhere `\u003Cslug>` is the team-member's canonical kebab-case name (e.g.\n`atlas`, `alice-chen`, `thrifty-otter`). The slug starts with a letter and\nuses only `[a-z0-9-]`. It is also the directory name under\n`context\u002Fteam-members\u002F\u003Cslug>\u002F`.\n\nFor `type=ai-agent`, the slug must be drawn from the team-manager name\npool (see **Name pool** below). For `type=human` and `type=service`, any\nvalid kebab-case slug is accepted.\n\n## TeamMember types\n\n| type         | Example                                     | When to create |\n|--------------|---------------------------------------------|----------------|\n| `human`      | Project owner, contributors, reviewers      | The first time the person is assigned work, makes a decision, or gets a Binding |\n| `ai-agent`   | Atlas, Aria, Kai (named personas)           | A persistent AI persona you want to grow over time — with memory, personality, and accumulated lessons |\n| `service`    | GitHub Actions, CI bots, deploy pipelines   | Automated agents that perform repeat actions and need a stable identity for audit trails |\n\nDo **not** promote every mention of a person into a TeamMember. And do\n**not** create a named AI persona for a one-off dispatch — use an\nephemeral role+seniority invocation instead.\n\n## Memory tiers\n\nEvery team-member directory contains six tier subdirectories under\n`context\u002Fteam-members\u002F\u003Cslug>\u002F`:\n\n| Tier       | What lives here | Cadence |\n|------------|-----------------|---------|\n| `working\u002F` | Task-local scratchpad, active session context | Ephemeral |\n| `journal\u002F` | Daily journal, episodic entries | Daily consolidation |\n| `knowledge\u002F` | Semantic knowledge, stable facts, domain notes | Weekly promotion from journal |\n| `skills\u002F`  | Procedural recipes, working style, patterns | As learned |\n| `relations\u002F` | Per-teammate relationship notes (one file per known teammate) | On interaction |\n| `lessons\u002F` | Reflected lessons; A-MemGuard-style consensus defense against memory poisoning | Per-task |\n\nPlus `private\u002F` (optional, developer-local, gitignored) and three\ntop-level files: `persona.md` (identity + personality prompt),\n`card.json` (A2A Agent Card), `team-member.md` (the entity itself).\n\nEvery memory file under a tier subdirectory must carry a frontmatter\nblock validated against `assets\u002Fmemory-file-header.schema.json` —\nrequired keys: `tier`, `source`, `sensitivity`, `created`.\n\n## Name pool & international naming policy\n\nThe name pool lives at `data\u002Fname-pool.yaml` (managed by this skill).\nIt contains three gender-bucket lists (feminine, masculine, neutral) of\ninternational, short, human-friendly first names — ~60 total. Names\nfrom the pool are the canonical identifier source for AI personas.\n\n**Reservation lifecycle**\n\n- `suggest_name(kind?)` — pick a random available name, optionally\n  filtered by gender bucket. Does not reserve.\n- `reserve_name(name, team_member_slug)` — mark a name taken by a\n  specific slug. Writes to `data\u002Fname-pool.yaml` atomically.\n- `release_name(name)` — remove the reservation (e.g. on team-member\n  deletion or re-export).\n- `list_available_names(kind?)` — enumerate non-reserved names.\n\nFor `create_team_member(type=\"ai-agent\", ...)`, the `slug` must\ncorrespond to a name currently in the pool. The tool auto-reserves on\nsuccess.\n\n## Consistency checks\n\n`check_consistency(slug)` returns findings for a single team-member;\n`check_all_consistency()` aggregates across every team-member.\n\nEach finding is `{severity, code, team_member, path, message}`. Codes:\n\n| Code | Severity | What it checks |\n|------|----------|----------------|\n| `team.drift.schema` | error | `team-member.yaml` frontmatter validates against `SCHEMA-team-member` |\n| `team.drift.tier_missing` | error | Expected subdirs (knowledge, journal, skills, relations, lessons) exist |\n| `team.drift.dangling_ref` | error | `default_role` resolves to an existing Role; `relationships[].with` resolves to known slugs\u002FIDs |\n| `team.name_collision` | error | No two active team-members share `name` or `slug` |\n| `team.name.off_pool` | warning | For `type=ai-agent`, the name must be in the name pool |\n| `team.drift.orphan_file` | warning | No files outside expected tier subdirs \u002F persona.md \u002F card.json \u002F team-member.md \u002F private\u002F |\n| `team.sensitivity.leak_risk` | warning | Memory files with `sensitivity: confidential` or `pii` must live under `private\u002F` |\n| `team.private.not_ignored` | warning | Each `private\u002F` directory is covered by a `.gitignore` rule |\n| `team.memory.bad_header` | warning | Memory files under tier subdirs carry required frontmatter keys |\n| `team.card.stale` | warning | `card.json` name\u002Frole\u002Fseniority match `team-member.yaml` |\n\n## Export \u002F import\n\n`export_team_member(slug, output_path?)` builds a tar.gz bundle at\n`\u003Cslug>-export-\u003CYYYY-MM-DD>.tar.gz` containing:\n\n- `persona.md`\n- `card.json` (A2A Agent Card, conforms to `assets\u002Fagent-card.schema.json`)\n- `team-member.md` (the entity)\n- `knowledge\u002F`, `skills\u002F`, `lessons\u002F`\n\nExcludes `journal\u002F`, `relations\u002F`, `private\u002F` by default. During the\nmemory-file walk, any file whose frontmatter declares\n`sensitivity: confidential` or `sensitivity: pii` is redacted (skipped).\n\n`import_team_member(tarball_path)` extracts to a temp dir, validates the\n`team-member.yaml` against `SCHEMA-team-member`, validates `card.json`\nagainst the A2A Agent Card schema, and checks that a `signature` field\nis present on the card (crypto verification is deferred — signature\nfield presence is a hand-off marker for future work). On success it\ncopies the tree into `context\u002Fteam-members\u002F\u003Cslug>\u002F` and logs\n`team_member.imported`.\n\n## Deprecation note\n\nThis skill **replaces** `actor-profile` (DEC-20260422_0233-SpryTulip).\nThe `ACTOR-\u003Cslug>` ID class and the `create_actor`\u002F`get_actor`\u002Fetc.\ntools are deprecated in v0.19.0. A migration on `context\u002Factors\u002F` runs\nin a separate phase; do not modify actor-profile from here.\n\n## Gotchas\n\n- **Creating a TeamMember for every mention.** Reserved for persistent\n  participants — the same rule as Actor. If they have no upcoming\n  WorkItem\u002FBinding\u002Fdecision, don't promote them.\n- **Drawing an AI persona name outside the pool.** `team.name.off_pool`\n  will warn. Use `suggest_name` or `list_available_names` first.\n- **Putting confidential info outside `private\u002F`.** `team.sensitivity.leak_risk`\n  catches it. The tier subdirectories are git-tracked; only `private\u002F`\n  is safe for developer-local sensitive context.\n- **Hand-editing `data\u002Fname-pool.yaml`.** Use `reserve_name`\u002F\n  `release_name` — the tool preserves the document frontmatter and\n  writes atomically.\n- **Forgetting `init_memory_tree` after create.** `create_team_member`\n  writes the entity but does not scaffold the tier subdirectories.\n  Call `init_memory_tree(slug)` right after (or pass\n  `init_memory=True` at create time, if implemented).\n- **Putting role+seniority dispatches here.** Ephemeral worker\n  invocations are NOT TeamMembers — they are resolved on the fly via\n  `task-router`. Only persistent personas live in `team-members\u002F`.\n\n## Full reference\n\n### MCP tool contract\n\nAll fifteen tools are served by `mcp\u002Fserver.py`. Their tool\nannotations follow the processkit convention checked by\n`release-audit.mcp_annotations`:\n\n| Tool | readOnlyHint | destructiveHint | idempotentHint | openWorldHint |\n|---|---|---|---|---|\n| `create_team_member` | false | false | false | false |\n| `get_team_member` | true | false | true | false |\n| `list_team_members` | true | false | true | false |\n| `update_team_member` | false | false | false | false |\n| `deactivate_team_member` | false | true | true | false |\n| `reactivate_team_member` | false | false | true | false |\n| `reserve_name` | false | false | false | false |\n| `release_name` | false | true | true | false |\n| `list_available_names` | true | false | true | false |\n| `suggest_name` | true | false | false | false |\n| `init_memory_tree` | false | false | true | false |\n| `export_team_member` | true | false | true | false |\n| `import_team_member` | false | false | false | false |\n| `check_consistency` | true | false | true | false |\n| `check_all_consistency` | true | false | true | false |\n\n`mcp\u002FSERVER.md` documents each tool's full input\u002Foutput schema\nincluding the JSON shape of `create_team_member` (slug, type,\ndefault_role, name, gender_kind, …) and the `Finding` schema for\nconsistency checks (`{severity, code, team_member, path, message}`).\n\n### Entity schema\n\nThe TeamMember frontmatter is validated against\n`SCHEMA-team-member` (registered in `context\u002Fschemas\u002F`). Required\nfields:\n\n| Field | Type | Notes |\n|---|---|---|\n| `id` | string | Must match `TEAMMEMBER-\u003Cslug>` |\n| `slug` | string | `[a-z][a-z0-9-]*`; equals directory name |\n| `name` | string | Display name; uniqueness enforced by `team.name_collision` |\n| `type` | enum | `human` \\| `ai-agent` \\| `service` |\n| `active` | bool | `false` after `deactivate_team_member` |\n| `default_role` | string \\| null | Resolved against the Role catalog by `team.drift.dangling_ref` |\n| `created` | ISO8601 | Stamped by `create_team_member` |\n| `relationships` | list | `[{with: \u003Cslug-or-id>, kind: \u003Cstring>, since: \u003CISO>}]`; references resolved by drift check |\n\nOptional fields used by templates: `card_path` (relative path to\n`card.json`), `persona_path` (relative path to `persona.md`),\n`memory_root` (relative path; defaults to the directory itself).\n\n### `data\u002Fname-pool.yaml` document format\n\n```yaml\n---\nschema: name-pool.v1\nmanaged_by: SKILL-team-manager\nupdated: \u003CISO8601>\n---\nfeminine:\n  - name: aria\n    reserved_by: TEAMMEMBER-aria   # null if available\n    reserved_at: \u003CISO8601>         # null if available\nmasculine:\n  - name: atlas\n    reserved_by: null\n    reserved_at: null\nneutral:\n  - name: kai\n    reserved_by: null\n    reserved_at: null\n```\n\nAtomic writes are produced by `scripts\u002Fname_pool.py` using a\nwrite-then-rename pattern; the document frontmatter (`schema`,\n`managed_by`, `updated`) is always preserved on round-trip. Hand\nedits silently break atomicity guarantees and are caught only by\nthe next consistency check.\n\n### Memory tier directory layout\n\n```\ncontext\u002Fteam-members\u002F\u003Cslug>\u002F\n├── persona.md                   # Identity + personality prompt\n├── card.json                    # A2A Agent Card (assets\u002Fagent-card.schema.json)\n├── team-member.md               # Entity file (frontmatter is the schema target)\n├── working\u002F                     # Ephemeral, task-local\n├── journal\u002F                     # Daily entries; consolidated weekly → knowledge\u002F\n├── knowledge\u002F                   # Stable semantic facts\n├── skills\u002F                      # Procedural recipes \u002F patterns\n├── relations\u002F                   # One file per known teammate slug\n├── lessons\u002F                     # A-MemGuard-style consensus lessons\n└── private\u002F                     # Optional, gitignored, developer-local\n```\n\nEvery memory file under `journal\u002F`, `knowledge\u002F`, `skills\u002F`,\n`relations\u002F`, `lessons\u002F`, or `private\u002F` carries a frontmatter block\nvalidated by `assets\u002Fmemory-file-header.schema.json`. Required keys:\n`tier`, `source`, `sensitivity`, `created`. Allowed `sensitivity`\nvalues: `public`, `internal`, `confidential`, `pii`. Files tagged\n`confidential` or `pii` outside `private\u002F` raise\n`team.sensitivity.leak_risk`.\n\n### Consistency check matrix (recap)\n\nThe 10 codes are split between hard errors (block release) and\nwarnings (advisory). The corresponding implementation lives in\n`scripts\u002Fconsistency.py`:\n\n| Code | Severity | Implementation function |\n|---|---|---|\n| `team.drift.schema` | error | `check_schema(slug)` |\n| `team.drift.tier_missing` | error | `check_tier_dirs(slug)` |\n| `team.drift.dangling_ref` | error | `check_refs(slug)` |\n| `team.name_collision` | error | `check_collisions()` (cross-slug) |\n| `team.name.off_pool` | warning | `check_pool_membership(slug)` |\n| `team.drift.orphan_file` | warning | `check_orphans(slug)` |\n| `team.sensitivity.leak_risk` | warning | `check_sensitivity(slug)` |\n| `team.private.not_ignored` | warning | `check_gitignore(slug)` |\n| `team.memory.bad_header` | warning | `check_memory_headers(slug)` |\n| `team.card.stale` | warning | `check_card(slug)` |\n\n`check_all_consistency()` aggregates over `list_team_members(active=true)`\nand returns `{summary: {errors, warnings}, findings: [...]}`. Use the\nsummary for CI gating; iterate findings for human-readable triage.\n\n### Export bundle layout\n\n```\n\u003Cslug>-export-\u003CYYYY-MM-DD>.tar.gz\n└── \u003Cslug>\u002F\n    ├── persona.md\n    ├── card.json\n    ├── team-member.md\n    ├── knowledge\u002F\n    ├── skills\u002F\n    └── lessons\u002F\n```\n\nExcludes `working\u002F`, `journal\u002F`, `relations\u002F`, `private\u002F`. During\nthe memory walk, any file with frontmatter `sensitivity:\nconfidential` or `sensitivity: pii` is skipped (redacted) regardless\nof which exported tier it lives in. Implementation:\n`scripts\u002Fexport_import.py::export_bundle()`.\n\n### Import contract\n\n`import_team_member(tarball_path)`:\n\n1. Extract to a temp dir (cleaned on failure).\n2. Validate `team-member.yaml` against `SCHEMA-team-member`.\n3. Validate `card.json` against `assets\u002Fagent-card.schema.json`.\n4. Require a `signature` field on the card (presence-only;\n   cryptographic verification is a future-work hand-off marker).\n5. Require the slug to be free locally (no collision with an\n   existing TeamMember entity or active name-pool reservation).\n6. Copy the tree to `context\u002Fteam-members\u002F\u003Cslug>\u002F`.\n7. For `type=ai-agent`, `reserve_name(name, slug)` is called as\n   part of step 6.\n8. Log `team_member.imported` via `event-log`.\n\nFailures at any step abort cleanly — the temp dir is removed and no\ntarget-side state is mutated.\n\n### Migration from `actor-profile` (DEC-20260422_0233-SpryTulip)\n\n`actor-profile` is deprecated as of v0.19.0; it is replaced by this\nskill. Do NOT modify `actor-profile` from here. The migration\nstrategy is:\n\n1. New write traffic targets `create_team_member` exclusively.\n2. Existing `ACTOR-\u003Cslug>` IDs continue to be readable via the\n   `actor-profile` MCP tools through the deprecation window.\n3. A separate migration phase (tracked by its own MIG-* entity)\n   converts each Actor under `context\u002Factors\u002F` into a TeamMember\n   under `context\u002Fteam-members\u002F\u003Cslug>\u002F`, preserving ID history via\n   `replaces` cross-references.\n4. After the migration completes, the `actor-profile` skill is\n   removed entirely; release-audit will then start flagging any\n   remaining `ACTOR-*` references as dangling.\n\nThe frontmatter `replaces: [SKILL-actor-profile]` records the\nrelationship; downstream skills should depend on `team-manager`,\nnot `actor-profile`, in their `uses:` blocks.\n\n### Referenced source files\n\n| Path | Purpose |\n|---|---|\n| `mcp\u002Fserver.py` | MCP server entry point (PEP 723) |\n| `mcp\u002FSERVER.md` | Per-tool input\u002Foutput schema reference |\n| `mcp\u002Fmcp-config.json` | Harness merge block for stdio launch |\n| `scripts\u002Fconsistency.py` | All 10 consistency check implementations |\n| `scripts\u002Fexport_import.py` | Tarball build + import validation pipeline |\n| `scripts\u002Fmemory_tree.py` | `init_memory_tree` scaffolding logic |\n| `scripts\u002Fname_pool.py` | Atomic reserve\u002Frelease\u002Flist\u002Fsuggest |\n| `scripts\u002Ftest_team_manager.py` | Pytest suite covering tools + scripts |\n| `data\u002Fname-pool.yaml` | The canonical name pool document |\n| `assets\u002Fagent-card.schema.json` | A2A Agent Card schema |\n| `assets\u002Fmemory-file-header.schema.json` | Memory file frontmatter schema |\n| `assets\u002Fteam-member-human.yaml` | Template for `type=human` |\n| `assets\u002Fteam-member-ai-agent.yaml` | Template for `type=ai-agent` |\n\n### Extension points\n\n- **New TeamMember type.** Add a new value to the `type` enum in\n  `SCHEMA-team-member`, ship a template under `assets\u002F`, extend the\n  name-pool policy if a curated slug source is needed, and update\n  the `team.name.off_pool` check to skip the new type if no curated\n  pool applies.\n- **New memory tier.** Add the directory to `init_memory_tree`,\n  document it in the Memory tiers table, and extend\n  `check_orphans` to recognise it. Decide whether export should\n  include or redact it.\n- **New consistency check.** Add a function in\n  `scripts\u002Fconsistency.py` returning `list[Finding]`, register it\n  in the dispatcher, and document the code + severity in the\n  matrix above. CI surfaces it on the next release-audit run.\n\n### Corner cases\n\n- **Slug collision on import.** Hard error; the importer never\n  overwrites an existing TeamMember directory or reservation. To\n  re-key a persona on import, rename the export tarball's inner\n  directory + `team-member.md` slug field before importing.\n- **Name reserved with no TeamMember.** Possible if a previous\n  `create_team_member` failed after `reserve_name` but before\n  entity write. Surfaced by `check_all_consistency()` as\n  `team.drift.dangling_ref` against the pool. Resolve by\n  `release_name` (manual triage, not auto-cleanup).\n- **`private\u002F` not gitignored.** `team.private.not_ignored`\n  warning. Add `context\u002Fteam-members\u002F\u003Cslug>\u002Fprivate\u002F` (or a glob\n  parent) to `.gitignore`. The skill never writes `.gitignore`\n  automatically — the warning is the action prompt.\n- **Card \u002F entity drift.** `team.card.stale` warning when\n  `card.json`'s `name` \u002F `role` \u002F `seniority` no longer match\n  `team-member.md` frontmatter. Re-export and re-import is the\n  blunt fix; targeted re-write of `card.json` is the surgical fix.\n- **Reactivating a deactivated member whose name was released.**\n  `reactivate_team_member` will re-reserve the name if it is still\n  free; if another persona has taken it, the call errors and the\n  caller must pick a new slug (or wait for the colliding persona\n  to be deactivated).\n","# Team Manager\n\n## Intro\n\nTeamMembers are the persistent participants in the project — named AI\npersonas, services, consultants, and explicitly managed human aliases —\nthat accumulate identity, personality, and tiered memory over time. This\nskill owns their lifecycle: creation, memory tree scaffolding, a curated\ninternational name pool, A2A-compatible Agent Cards, export\u002Fimport\nbundles, and consistency checks.\n\nUnlike ephemeral worker invocations (role+seniority dispatches resolved on\nthe fly and never persisted), TeamMembers live as directory trees on disk:\n`persona.md` + `card.json` + `team-member.md` (entity) + tiered memory\nsubdirectories + optional `private\u002F` (gitignored).\n\nHuman TeamMembers are privacy-sensitive. Do not auto-create them from\nambient host, git, or harness identity. `create_team_member(type=\"human\",\n...)` stores only the slug as the committed display name by default and\ndrops `email`\u002F`handle`; pass `allow_committed_pii=true` only after the\nhuman explicitly agrees that their name or contact details may live in\nrepo-visible files. Store local-only human details under `private\u002F`, not\nin `team-member.md`.\n\n> **MCP server.** This skill ships a self-contained MCP server at\n> `mcp\u002Fserver.py` (PEP 723 script — requires `uv` and Python ≥ 3.10 on\n> PATH). Agent harnesses reach its tools by reading a single MCP config\n> file at startup, so the contents of `mcp\u002Fmcp-config.json` must be merged\n> into the harness's MCP config and placed at the harness-specific path\n> before this skill is usable.\n\n## Overview\n\n### Invocation surface\n\nThis skill is not a CLI; it is reached entirely via MCP tools served\nby `mcp\u002Fserver.py`. Tools cluster into six capability groups\n(lifecycle, session identity, name pool, memory scaffolding,\nexport\u002Fimport, consistency). Every write tool logs an `event-log`\nentry and updates the SQLite index automatically — callers do not log\nor reindex by hand. Read tools (`get_team_member`,\n`list_team_members`, `get_active_interlocutor`,\n`get_interlocutor_runtime_binding`, `list_available_names`,\n`suggest_name`, `get_team_member_runtime`, `list_team_member_runtimes`,\n`check_consistency`) are side-effect-free and safe to call from any\nagent context.\n\n### Capability groups\n\n| Group | Tools | Use when |\n|---|---|---|\n| Lifecycle | `create_team_member`, `get_team_member`, `list_team_members`, `update_team_member`, `deactivate_team_member`, `reactivate_team_member` | Adding, editing, retiring or restoring a persistent team-member entity |\n| Session identity | `get_active_interlocutor`, `get_interlocutor_runtime_binding`, `set_active_interlocutor` | Showing which persistent TeamMember, if any, is the current harness interlocutor, plus the resolved runtime binding when available |\n| Runtime launch | `launch_team_member`, `launch_workitem_assignee`, `get_team_member_runtime`, `list_team_member_runtimes`, `stop_team_member_runtime` | Creating and tracking a harness-consumable TeamMember runtime request with explicit model, status, and write policy |\n| Name pool | `reserve_name`, `release_name`, `list_available_names`, `suggest_name` | Choosing or rotating an AI persona name; the pool is the canonical source of truth for AI-agent slugs |\n| Memory scaffolding | `init_memory_tree` | Creating the six tier subdirectories + frontmatter-validated headers under `context\u002Fteam-members\u002F\u003Cslug>\u002F` |\n| Export \u002F import | `export_team_member`, `export_claude_subagent`, `export_claude_subagents`, `import_team_member` | Moving a TeamMember between projects or generating provider-specific adapter files; tarball export redacts confidential\u002FPII memory |\n| Consistency | `check_consistency`, `check_all_consistency` | Pre-release health-check; CI validation; ad-hoc audits, including human committed-PII warnings |\n\n### Common workflows\n\n**Adding a human contributor.** Prefer an ephemeral Actor\u002Freference unless\nthe project really needs a persistent human TeamMember. When persistence\nis needed, use a privacy-preserving slug such as `human-user` or\n`project-owner`: `create_team_member(type=\"human\", name=\"human-user\",\nslug=\"human-user\", ...)` → `init_memory_tree(slug)`. No name-pool\ninteraction; humans use any valid kebab-case slug. Real names, emails,\nand handles require `allow_committed_pii=true` and explicit human consent.\n\n**Adding a named AI persona.** `suggest_name(kind=\"neutral\")` →\n`create_team_member(type=\"ai-agent\", slug=\u003Cname>, ...)` (auto-reserves\nthe name) → `init_memory_tree(slug)` → write `persona.md` and the\nA2A-compliant `card.json`. The name is now drawn from the pool and\nmust remain in `data\u002Fname-pool.yaml` until released.\n\n**Retiring a persona.** `deactivate_team_member(slug, reason)` (sets\n`active: false`; preserves history) → optionally\n`release_name(name)` to return the slug to the pool. To bring the\npersona back, call `reactivate_team_member(slug)`.\n\n**Migrating a persona between projects.**\n`export_team_member(slug)` → ship the tarball → at the destination\n`import_team_member(tarball_path)`. Schema and Agent Card are\nre-validated on import; signature presence is required (crypto\nverification is deferred).\n\n**Showing the active interlocutor.**\n`set_active_interlocutor(id)` writes\n`context\u002Fteam\u002Fsession-identity.json` with the TeamMember that should be\npresented as the harness's main speaker. `get_active_interlocutor()`\nreturns a compact label and speaker prefix for status briefings and\nsession-start UI. This is session identity, not lifecycle state:\n`TeamMember.spec.active` continues to mean whether a TeamMember is\neligible to participate at all.\n\nUse `get_interlocutor_runtime_binding(observed_model?,\nobserved_effort?)` when the caller can supply the harness's actual\nruntime. The tool returns the provider-neutral TeamMember identity, the\nmodel-recommender candidate after local provider gates, the negotiated\nharness mode, and an informational mismatch report. processkit cannot\nhot-swap an already running primary harness session; unsupported\nharnesses therefore fall back to identity-only behavior while still\ngrowing memory on the configured TeamMember.\n\n**Launching TeamMember runtimes.**\n`launch_team_member(id, harness?, ...)` creates a runtime record under\n`context\u002Fteam\u002Fruntime-sessions.json` and returns a harness-specific\ndispatch payload. `launch_workitem_assignee(workitem_id, ...)` resolves\nthe WorkItem's `TEAMMEMBER-*` assignee first and then records the same\nruntime contract. Runtime state is separate from WorkItem state:\n`queued`, `running`, `failed`, and `stopped` describe the harness\nsession, not the work lifecycle.\n\nThe MCP server does not spawn host harness processes directly. For\nClaude, the dispatch payload names the `subagent_type` and refreshes the\n`.claude\u002Fagents\u002F\u003Cslug>.md` adapter by default. For Codex, Aider, and\nOpenCode, the payload carries the TeamMember identity, resolved model,\neffort, write scope, and MCP\u002Fcontext-write policy for the outer harness\nadapter to consume. Use `list_team_member_runtimes(active_only=true)`\nand `stop_team_member_runtime(handle)` to report and update status.\n\n**Exporting Claude Code subagents.**\n`export_claude_subagent(slug)` writes one generated adapter to\n`.claude\u002Fagents\u002F\u003Cslug>.md`; `export_claude_subagents()` exports all\nactive non-human TeamMembers by default. The generated file uses\nClaude Code's Markdown + YAML frontmatter format, sets `model:\ninherit`, and omits `tools` so Claude Code inherits the main session's\nmodel and permissions. Pass `model_policy=\"resolved\"` to emit a Claude\nmodel and effort only when processkit routing resolves to an Anthropic\ncandidate. TeamMember remains the canonical, provider-neutral identity;\n`.claude\u002Fagents\u002F*.md` is only an adapter.\n\n**Pre-release audit.** `check_all_consistency()` → triage findings by\nseverity. Errors block release; warnings are advisory. Treat\n`team.privacy.committed_pii` as release-blocking for derived-project\ntemplates and installer output: rotate the TeamMember to an alias or move\nthe personal data into `private\u002F`.\n\n### Mental model\n\nA TeamMember is **a directory tree on disk**, not a row in a database.\nThe entity file (`team-member.md` \u002F `team-member.yaml` frontmatter)\nis the SQLite-indexed object; `persona.md`, `card.json`, and the six\ntier subdirectories live alongside it as durable, human-editable\ncontext. Humans\u002Fservices have no name-pool requirement; AI personas\ndo. TeamMembers are distinct from **ephemeral worker invocations** —\nthose are resolved on the fly by `task-router` against role+seniority\nand never persist as TeamMembers.\n\n### Pointers to the full reference\n\nThe `## Full reference` section below covers the MCP tool contract,\nthe SQLite-index schema fields, the `data\u002Fname-pool.yaml` document\nformat, the consistency check codes in greater detail, the export\nbundle layout, the migration path from `actor-profile`\n(DEC-20260422_0233-SpryTulip), and the corner cases for memory\nsensitivity tagging. Read it before extending the skill or changing\nany tool's contract.\n\n## ID convention\n\nTeamMember IDs follow a single shape:\n\n    TEAMMEMBER-\u003Cslug>\n\nwhere `\u003Cslug>` is the team-member's canonical kebab-case name (e.g.\n`atlas`, `alice-chen`, `thrifty-otter`). The slug starts with a letter and\nuses only `[a-z0-9-]`. It is also the directory name under\n`context\u002Fteam-members\u002F\u003Cslug>\u002F`.\n\nFor `type=ai-agent`, the slug must be drawn from the team-manager name\npool (see **Name pool** below). For `type=human` and `type=service`, any\nvalid kebab-case slug is accepted.\n\n## TeamMember types\n\n| type         | Example                                     | When to create |\n|--------------|---------------------------------------------|----------------|\n| `human`      | Project owner, contributors, reviewers      | The first time the person is assigned work, makes a decision, or gets a Binding |\n| `ai-agent`   | Atlas, Aria, Kai (named personas)           | A persistent AI persona you want to grow over time — with memory, personality, and accumulated lessons |\n| `service`    | GitHub Actions, CI bots, deploy pipelines   | Automated agents that perform repeat actions and need a stable identity for audit trails |\n\nDo **not** promote every mention of a person into a TeamMember. And do\n**not** create a named AI persona for a one-off dispatch — use an\nephemeral role+seniority invocation instead.\n\n## Memory tiers\n\nEvery team-member directory contains six tier subdirectories under\n`context\u002Fteam-members\u002F\u003Cslug>\u002F`:\n\n| Tier       | What lives here | Cadence |\n|------------|-----------------|---------|\n| `working\u002F` | Task-local scratchpad, active session context | Ephemeral |\n| `journal\u002F` | Daily journal, episodic entries | Daily consolidation |\n| `knowledge\u002F` | Semantic knowledge, stable facts, domain notes | Weekly promotion from journal |\n| `skills\u002F`  | Procedural recipes, working style, patterns | As learned |\n| `relations\u002F` | Per-teammate relationship notes (one file per known teammate) | On interaction |\n| `lessons\u002F` | Reflected lessons; A-MemGuard-style consensus defense against memory poisoning | Per-task |\n\nPlus `private\u002F` (optional, developer-local, gitignored) and three\ntop-level files: `persona.md` (identity + personality prompt),\n`card.json` (A2A Agent Card), `team-member.md` (the entity itself).\n\nEvery memory file under a tier subdirectory must carry a frontmatter\nblock validated against `assets\u002Fmemory-file-header.schema.json` —\nrequired keys: `tier`, `source`, `sensitivity`, `created`.\n\n## Name pool & international naming policy\n\nThe name pool lives at `data\u002Fname-pool.yaml` (managed by this skill).\nIt contains three gender-bucket lists (feminine, masculine, neutral) of\ninternational, short, human-friendly first names — ~60 total. Names\nfrom the pool are the canonical identifier source for AI personas.\n\n**Reservation lifecycle**\n\n- `suggest_name(kind?)` — pick a random available name, optionally\n  filtered by gender bucket. Does not reserve.\n- `reserve_name(name, team_member_slug)` — mark a name taken by a\n  specific slug. Writes to `data\u002Fname-pool.yaml` atomically.\n- `release_name(name)` — remove the reservation (e.g. on team-member\n  deletion or re-export).\n- `list_available_names(kind?)` — enumerate non-reserved names.\n\nFor `create_team_member(type=\"ai-agent\", ...)`, the `slug` must\ncorrespond to a name currently in the pool. The tool auto-reserves on\nsuccess.\n\n## Consistency checks\n\n`check_consistency(slug)` returns findings for a single team-member;\n`check_all_consistency()` aggregates across every team-member.\n\nEach finding is `{severity, code, team_member, path, message}`. Codes:\n\n| Code | Severity | What it checks |\n|------|----------|----------------|\n| `team.drift.schema` | error | `team-member.yaml` frontmatter validates against `SCHEMA-team-member` |\n| `team.drift.tier_missing` | error | Expected subdirs (knowledge, journal, skills, relations, lessons) exist |\n| `team.drift.dangling_ref` | error | `default_role` resolves to an existing Role; `relationships[].with` resolves to known slugs\u002FIDs |\n| `team.name_collision` | error | No two active team-members share `name` or `slug` |\n| `team.name.off_pool` | warning | For `type=ai-agent`, the name must be in the name pool |\n| `team.drift.orphan_file` | warning | No files outside expected tier subdirs \u002F persona.md \u002F card.json \u002F team-member.md \u002F private\u002F |\n| `team.sensitivity.leak_risk` | warning | Memory files with `sensitivity: confidential` or `pii` must live under `private\u002F` |\n| `team.private.not_ignored` | warning | Each `private\u002F` directory is covered by a `.gitignore` rule |\n| `team.memory.bad_header` | warning | Memory files under tier subdirs carry required frontmatter keys |\n| `team.card.stale` | warning | `card.json` name\u002Frole\u002Fseniority match `team-member.yaml` |\n\n## Export \u002F import\n\n`export_team_member(slug, output_path?)` builds a tar.gz bundle at\n`\u003Cslug>-export-\u003CYYYY-MM-DD>.tar.gz` containing:\n\n- `persona.md`\n- `card.json` (A2A Agent Card, conforms to `assets\u002Fagent-card.schema.json`)\n- `team-member.md` (the entity)\n- `knowledge\u002F`, `skills\u002F`, `lessons\u002F`\n\nExcludes `journal\u002F`, `relations\u002F`, `private\u002F` by default. During the\nmemory-file walk, any file whose frontmatter declares\n`sensitivity: confidential` or `sensitivity: pii` is redacted (skipped).\n\n`import_team_member(tarball_path)` extracts to a temp dir, validates the\n`team-member.yaml` against `SCHEMA-team-member`, validates `card.json`\nagainst the A2A Agent Card schema, and checks that a `signature` field\nis present on the card (crypto verification is deferred — signature\nfield presence is a hand-off marker for future work). On success it\ncopies the tree into `context\u002Fteam-members\u002F\u003Cslug>\u002F` and logs\n`team_member.imported`.\n\n## Deprecation note\n\nThis skill **replaces** `actor-profile` (DEC-20260422_0233-SpryTulip).\nThe `ACTOR-\u003Cslug>` ID class and the `create_actor`\u002F`get_actor`\u002Fetc.\ntools are deprecated in v0.19.0. A migration on `context\u002Factors\u002F` runs\nin a separate phase; do not modify actor-profile from here.\n\n## Gotchas\n\n- **Creating a TeamMember for every mention.** Reserved for persistent\n  participants — the same rule as Actor. If they have no upcoming\n  WorkItem\u002FBinding\u002Fdecision, don't promote them.\n- **Drawing an AI persona name outside the pool.** `team.name.off_pool`\n  will warn. Use `suggest_name` or `list_available_names` first.\n- **Putting confidential info outside `private\u002F`.** `team.sensitivity.leak_risk`\n  catches it. The tier subdirectories are git-tracked; only `private\u002F`\n  is safe for developer-local sensitive context.\n- **Hand-editing `data\u002Fname-pool.yaml`.** Use `reserve_name`\u002F\n  `release_name` — the tool preserves the document frontmatter and\n  writes atomically.\n- **Forgetting `init_memory_tree` after create.** `create_team_member`\n  writes the entity but does not scaffold the tier subdirectories.\n  Call `init_memory_tree(slug)` right after (or pass\n  `init_memory=True` at create time, if implemented).\n- **Putting role+seniority dispatches here.** Ephemeral worker\n  invocations are NOT TeamMembers — they are resolved on the fly via\n  `task-router`. Only persistent personas live in `team-members\u002F`.\n\n## Full reference\n\n### MCP tool contract\n\nAll fifteen tools are served by `mcp\u002Fserver.py`. Their tool\nannotations follow the processkit convention checked by\n`release-audit.mcp_annotations`:\n\n| Tool | readOnlyHint | destructiveHint | idempotentHint | openWorldHint |\n|---|---|---|---|---|\n| `create_team_member` | false | false | false | false |\n| `get_team_member` | true | false | true | false |\n| `list_team_members` | true | false | true | false |\n| `update_team_member` | false | false | false | false |\n| `deactivate_team_member` | false | true | true | false |\n| `reactivate_team_member` | false | false | true | false |\n| `reserve_name` | false | false | false | false |\n| `release_name` | false | true | true | false |\n| `list_available_names` | true | false | true | false |\n| `suggest_name` | true | false | false | false |\n| `init_memory_tree` | false | false | true | false |\n| `export_team_member` | true | false | true | false |\n| `import_team_member` | false | false | false | false |\n| `check_consistency` | true | false | true | false |\n| `check_all_consistency` | true | false | true | false |\n\n`mcp\u002FSERVER.md` documents each tool's full input\u002Foutput schema\nincluding the JSON shape of `create_team_member` (slug, type,\ndefault_role, name, gender_kind, …) and the `Finding` schema for\nconsistency checks (`{severity, code, team_member, path, message}`).\n\n### Entity schema\n\nThe TeamMember frontmatter is validated against\n`SCHEMA-team-member` (registered in `context\u002Fschemas\u002F`). Required\nfields:\n\n| Field | Type | Notes |\n|---|---|---|\n| `id` | string | Must match `TEAMMEMBER-\u003Cslug>` |\n| `slug` | string | `[a-z][a-z0-9-]*`; equals directory name |\n| `name` | string | Display name; uniqueness enforced by `team.name_collision` |\n| `type` | enum | `human` \\| `ai-agent` \\| `service` |\n| `active` | bool | `false` after `deactivate_team_member` |\n| `default_role` | string \\| null | Resolved against the Role catalog by `team.drift.dangling_ref` |\n| `created` | ISO8601 | Stamped by `create_team_member` |\n| `relationships` | list | `[{with: \u003Cslug-or-id>, kind: \u003Cstring>, since: \u003CISO>}]`; references resolved by drift check |\n\nOptional fields used by templates: `card_path` (relative path to\n`card.json`), `persona_path` (relative path to `persona.md`),\n`memory_root` (relative path; defaults to the directory itself).\n\n### `data\u002Fname-pool.yaml` document format\n\n```yaml\n---\nschema: name-pool.v1\nmanaged_by: SKILL-team-manager\nupdated: \u003CISO8601>\n---\nfeminine:\n  - name: aria\n    reserved_by: TEAMMEMBER-aria   # null if available\n    reserved_at: \u003CISO8601>         # null if available\nmasculine:\n  - name: atlas\n    reserved_by: null\n    reserved_at: null\nneutral:\n  - name: kai\n    reserved_by: null\n    reserved_at: null\n```\n\nAtomic writes are produced by `scripts\u002Fname_pool.py` using a\nwrite-then-rename pattern; the document frontmatter (`schema`,\n`managed_by`, `updated`) is always preserved on round-trip. Hand\nedits silently break atomicity guarantees and are caught only by\nthe next consistency check.\n\n### Memory tier directory layout\n\n```\ncontext\u002Fteam-members\u002F\u003Cslug>\u002F\n├── persona.md                   # Identity + personality prompt\n├── card.json                    # A2A Agent Card (assets\u002Fagent-card.schema.json)\n├── team-member.md               # Entity file (frontmatter is the schema target)\n├── working\u002F                     # Ephemeral, task-local\n├── journal\u002F                     # Daily entries; consolidated weekly → knowledge\u002F\n├── knowledge\u002F                   # Stable semantic facts\n├── skills\u002F                      # Procedural recipes \u002F patterns\n├── relations\u002F                   # One file per known teammate slug\n├── lessons\u002F                     # A-MemGuard-style consensus lessons\n└── private\u002F                     # Optional, gitignored, developer-local\n```\n\nEvery memory file under `journal\u002F`, `knowledge\u002F`, `skills\u002F`,\n`relations\u002F`, `lessons\u002F`, or `private\u002F` carries a frontmatter block\nvalidated by `assets\u002Fmemory-file-header.schema.json`. Required keys:\n`tier`, `source`, `sensitivity`, `created`. Allowed `sensitivity`\nvalues: `public`, `internal`, `confidential`, `pii`. Files tagged\n`confidential` or `pii` outside `private\u002F` raise\n`team.sensitivity.leak_risk`.\n\n### Consistency check matrix (recap)\n\nThe 10 codes are split between hard errors (block release) and\nwarnings (advisory). The corresponding implementation lives in\n`scripts\u002Fconsistency.py`:\n\n| Code | Severity | Implementation function |\n|---|---|---|\n| `team.drift.schema` | error | `check_schema(slug)` |\n| `team.drift.tier_missing` | error | `check_tier_dirs(slug)` |\n| `team.drift.dangling_ref` | error | `check_refs(slug)` |\n| `team.name_collision` | error | `check_collisions()` (cross-slug) |\n| `team.name.off_pool` | warning | `check_pool_membership(slug)` |\n| `team.drift.orphan_file` | warning | `check_orphans(slug)` |\n| `team.sensitivity.leak_risk` | warning | `check_sensitivity(slug)` |\n| `team.private.not_ignored` | warning | `check_gitignore(slug)` |\n| `team.memory.bad_header` | warning | `check_memory_headers(slug)` |\n| `team.card.stale` | warning | `check_card(slug)` |\n\n`check_all_consistency()` aggregates over `list_team_members(active=true)`\nand returns `{summary: {errors, warnings}, findings: [...]}`. Use the\nsummary for CI gating; iterate findings for human-readable triage.\n\n### Export bundle layout\n\n```\n\u003Cslug>-export-\u003CYYYY-MM-DD>.tar.gz\n└── \u003Cslug>\u002F\n    ├── persona.md\n    ├── card.json\n    ├── team-member.md\n    ├── knowledge\u002F\n    ├── skills\u002F\n    └── lessons\u002F\n```\n\nExcludes `working\u002F`, `journal\u002F`, `relations\u002F`, `private\u002F`. During\nthe memory walk, any file with frontmatter `sensitivity:\nconfidential` or `sensitivity: pii` is skipped (redacted) regardless\nof which exported tier it lives in. Implementation:\n`scripts\u002Fexport_import.py::export_bundle()`.\n\n### Import contract\n\n`import_team_member(tarball_path)`:\n\n1. Extract to a temp dir (cleaned on failure).\n2. Validate `team-member.yaml` against `SCHEMA-team-member`.\n3. Validate `card.json` against `assets\u002Fagent-card.schema.json`.\n4. Require a `signature` field on the card (presence-only;\n   cryptographic verification is a future-work hand-off marker).\n5. Require the slug to be free locally (no collision with an\n   existing TeamMember entity or active name-pool reservation).\n6. Copy the tree to `context\u002Fteam-members\u002F\u003Cslug>\u002F`.\n7. For `type=ai-agent`, `reserve_name(name, slug)` is called as\n   part of step 6.\n8. Log `team_member.imported` via `event-log`.\n\nFailures at any step abort cleanly — the temp dir is removed and no\ntarget-side state is mutated.\n\n### Migration from `actor-profile` (DEC-20260422_0233-SpryTulip)\n\n`actor-profile` is deprecated as of v0.19.0; it is replaced by this\nskill. Do NOT modify `actor-profile` from here. The migration\nstrategy is:\n\n1. New write traffic targets `create_team_member` exclusively.\n2. Existing `ACTOR-\u003Cslug>` IDs continue to be readable via the\n   `actor-profile` MCP tools through the deprecation window.\n3. A separate migration phase (tracked by its own MIG-* entity)\n   converts each Actor under `context\u002Factors\u002F` into a TeamMember\n   under `context\u002Fteam-members\u002F\u003Cslug>\u002F`, preserving ID history via\n   `replaces` cross-references.\n4. After the migration completes, the `actor-profile` skill is\n   removed entirely; release-audit will then start flagging any\n   remaining `ACTOR-*` references as dangling.\n\nThe frontmatter `replaces: [SKILL-actor-profile]` records the\nrelationship; downstream skills should depend on `team-manager`,\nnot `actor-profile`, in their `uses:` blocks.\n\n### Referenced source files\n\n| Path | Purpose |\n|---|---|\n| `mcp\u002Fserver.py` | MCP server entry point (PEP 723) |\n| `mcp\u002FSERVER.md` | Per-tool input\u002Foutput schema reference |\n| `mcp\u002Fmcp-config.json` | Harness merge block for stdio launch |\n| `scripts\u002Fconsistency.py` | All 10 consistency check implementations |\n| `scripts\u002Fexport_import.py` | Tarball build + import validation pipeline |\n| `scripts\u002Fmemory_tree.py` | `init_memory_tree` scaffolding logic |\n| `scripts\u002Fname_pool.py` | Atomic reserve\u002Frelease\u002Flist\u002Fsuggest |\n| `scripts\u002Ftest_team_manager.py` | Pytest suite covering tools + scripts |\n| `data\u002Fname-pool.yaml` | The canonical name pool document |\n| `assets\u002Fagent-card.schema.json` | A2A Agent Card schema |\n| `assets\u002Fmemory-file-header.schema.json` | Memory file frontmatter schema |\n| `assets\u002Fteam-member-human.yaml` | Template for `type=human` |\n| `assets\u002Fteam-member-ai-agent.yaml` | Template for `type=ai-agent` |\n\n### Extension points\n\n- **New TeamMember type.** Add a new value to the `type` enum in\n  `SCHEMA-team-member`, ship a template under `assets\u002F`, extend the\n  name-pool policy if a curated slug source is needed, and update\n  the `team.name.off_pool` check to skip the new type if no curated\n  pool applies.\n- **New memory tier.** Add the directory to `init_memory_tree`,\n  document it in the Memory tiers table, and extend\n  `check_orphans` to recognise it. Decide whether export should\n  include or redact it.\n- **New consistency check.** Add a function in\n  `scripts\u002Fconsistency.py` returning `list[Finding]`, register it\n  in the dispatcher, and document the code + severity in the\n  matrix above. CI surfaces it on the next release-audit run.\n\n### Corner cases\n\n- **Slug collision on import.** Hard error; the importer never\n  overwrites an existing TeamMember directory or reservation. To\n  re-key a persona on import, rename the export tarball's inner\n  directory + `team-member.md` slug field before importing.\n- **Name reserved with no TeamMember.** Possible if a previous\n  `create_team_member` failed after `reserve_name` but before\n  entity write. Surfaced by `check_all_consistency()` as\n  `team.drift.dangling_ref` against the pool. Resolve by\n  `release_name` (manual triage, not auto-cleanup).\n- **`private\u002F` not gitignored.** `team.private.not_ignored`\n  warning. Add `context\u002Fteam-members\u002F\u003Cslug>\u002Fprivate\u002F` (or a glob\n  parent) to `.gitignore`. The skill never writes `.gitignore`\n  automatically — the warning is the action prompt.\n- **Card \u002F entity drift.** `team.card.stale` warning when\n  `card.json`'s `name` \u002F `role` \u002F `seniority` no longer match\n  `team-member.md` frontmatter. Re-export and re-import is the\n  blunt fix; targeted re-write of `card.json` is the surgical fix.\n- **Reactivating a deactivated member whose name was released.**\n  `reactivate_team_member` will re-reserve the name if it is still\n  free; if another persona has taken it, the call errors and the\n  caller must pick a new slug (or wait for the colliding persona\n  to be deactivated).\n",{"name":5,"metadata":49,"description":98},{"processkit":50},{"id":51,"uses":52,"layer":65,"created":66,"version":67,"category":68,"provides":69,"replaces":95,"apiVersion":97},"SKILL-team-manager",[53,56,59,62],{"skill":54,"purpose":55},"event-log","Log events to keep the audit trail accurate after every write.",{"skill":57,"purpose":58},"index-management","Query existing entities and keep the SQLite index fresh after writes.",{"skill":60,"purpose":61},"id-management","Allocate unique entity identifiers via central ID generation.",{"skill":63,"purpose":64},"role-management","Resolve default_role references against the Role catalog.",1,1776816000000,"1.0.0","processkit",{"mcp_tools":70,"templates":90,"primitives":93},[71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89],"create_team_member","get_team_member","list_team_members","get_active_interlocutor","set_active_interlocutor","update_team_member","deactivate_team_member","reactivate_team_member","reserve_name","release_name","list_available_names","suggest_name","init_memory_tree","export_team_member","export_claude_subagent","export_claude_subagents","import_team_member","check_consistency","check_all_consistency",[91,92],"team-member-human","team-member-ai-agent",[94],"TeamMember",[96],"SKILL-actor-profile","processkit.projectious.work\u002Fv2","Create and maintain TeamMember entities — humans and named AI personas that participate in the project with persistent identity, personality, and tiered memory. Use when adding a new persistent team-member (human or named AI), managing the name pool, scaffolding memory trees, or exporting\u002Fimporting a team-member bundle. Replaces the deprecated actor-profile skill (DEC-20260422_0233-SpryTulip).\n","main","7231f100e266e30d2c1699ecde5c00602471dc3c","src\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit","https:\u002F\u002Fprojectious-work.github.io\u002Fprocesskit\u002F","https:\u002F\u002Fgithub.com\u002Fprojectious-work","2026-11-11 23:24:18",{"source":8,"skillName":5,"commands":107,"prompt":117,"zipAvailable":118},[108,111,114],{"tool":109,"command":110},"npx","npx skills add projectious-work\u002Fprocesskit --skill team-manager",{"tool":112,"command":113},"bunx","bunx skills add projectious-work\u002Fprocesskit --skill team-manager",{"tool":115,"command":116},"pnpm","pnpm dlx skills add projectious-work\u002Fprocesskit --skill team-manager","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002FSKILL.md",true,[120,126,130,134,138,144,148,152,158,162,166,170,174,178],{"path":121,"resourceType":122,"resourceTypeDesc":123,"sizeBytes":124,"sourceHtmlUrl":125},"assets\u002Fagent-card.schema.json","ASSET","资产文件",869,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fassets\u002Fagent-card.schema.json",{"path":127,"resourceType":122,"resourceTypeDesc":123,"sizeBytes":128,"sourceHtmlUrl":129},"assets\u002Fmemory-file-header.schema.json",897,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fassets\u002Fmemory-file-header.schema.json",{"path":131,"resourceType":122,"resourceTypeDesc":123,"sizeBytes":132,"sourceHtmlUrl":133},"assets\u002Fteam-member-ai-agent.yaml",1047,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fassets\u002Fteam-member-ai-agent.yaml",{"path":135,"resourceType":122,"resourceTypeDesc":123,"sizeBytes":136,"sourceHtmlUrl":137},"assets\u002Fteam-member-human.yaml",1021,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fassets\u002Fteam-member-human.yaml",{"path":139,"resourceType":140,"resourceTypeDesc":141,"sizeBytes":142,"sourceHtmlUrl":143},"data\u002Fname-pool.yaml","REFERENCE","参考文档",1249,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fdata\u002Fname-pool.yaml",{"path":145,"resourceType":140,"resourceTypeDesc":141,"sizeBytes":146,"sourceHtmlUrl":147},"mcp\u002Fmcp-config.json",207,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fmcp\u002Fmcp-config.json",{"path":149,"resourceType":140,"resourceTypeDesc":141,"sizeBytes":150,"sourceHtmlUrl":151},"mcp\u002FSERVER.md",5472,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fmcp\u002FSERVER.md",{"path":153,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":156,"sourceHtmlUrl":157},"mcp\u002Fserver.py","SCRIPT","脚本",109989,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fmcp\u002Fserver.py",{"path":159,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":160,"sourceHtmlUrl":161},"scripts\u002Fapply_migration_2139.py",14459,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Fapply_migration_2139.py",{"path":163,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":164,"sourceHtmlUrl":165},"scripts\u002Fconsistency.py",15826,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Fconsistency.py",{"path":167,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":168,"sourceHtmlUrl":169},"scripts\u002Fexport_import.py",7325,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Fexport_import.py",{"path":171,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":172,"sourceHtmlUrl":173},"scripts\u002Fmemory_tree.py",3676,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Fmemory_tree.py",{"path":175,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":176,"sourceHtmlUrl":177},"scripts\u002Fname_pool.py",4754,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Fname_pool.py",{"path":179,"resourceType":154,"resourceTypeDesc":155,"sizeBytes":180,"sourceHtmlUrl":181},"scripts\u002Ftest_team_manager.py",89941,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fteam-manager\u002Fscripts\u002Ftest_team_manager.py",[183,190,197,203,224,244],{"skillNo":184,"name":185,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":186,"categories":187,"keywords":188,"publishedAt":189},"01M7665G2ED3624R3X93304265","standard-ui-design","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fstandard-ui-design\u002FSKILL.md",[],[],"2026-11-12 13:55:43",{"skillNo":191,"name":192,"nameZh":6,"description":193,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":194,"categories":195,"keywords":196,"publishedAt":189},"01M76651QTC6W0HJ4AFW2MMR3V","standard-agent-use","Company standard for working with AI agents. Use when deciding what to delegate to agents, capturing open work, keeping agent memory or knowledge notes, improving an agent workflow, or introducing agent work in a team or for a client.","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fstandard-agent-use\u002FSKILL.md",[],[],{"skillNo":198,"name":199,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":200,"categories":201,"keywords":202,"publishedAt":189},"01M76650NN4T8P241021NBR64S","standard-agent-building","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fstandard-agent-building\u002FSKILL.md",[],[],{"skillNo":204,"name":205,"nameZh":6,"description":206,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":207,"categories":208,"keywords":210,"publishedAt":44},"01M6X5MTC89B4ZGXW3JVSQCPF8","skill-management","Manage Skill package manifests as v1 ontology entities.","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fskill-management\u002FSKILL.md",[209],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[211,212,213,214,42,215,38,216,29,217,218,219,205,220,221,222,223],"management","skill management","ontology","v1","as","transition","manage","update","alpha","list","package","skill package","manifests",{"skillNo":225,"name":226,"nameZh":6,"description":227,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":228,"categories":229,"keywords":231,"publishedAt":44},"01M6X5MCV74NHW5MEDNHZRF2C8","schema-management","Manage generated processkit schema contracts through deterministic MCP rebuild and inspection workflows.","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fschema-management\u002FSKILL.md",[230],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[232,233,211,234,235,236,237,238,239,240,241,242,38,29,217,219,243],"schema","schema management","contract","rebuild","contracts","workflows","wed","generated","MCP","validation","deterministic","mode",{"skillNo":245,"name":246,"nameZh":6,"description":247,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":248,"categories":249,"keywords":250,"publishedAt":44},"01M6X5M5G6YJ96AMZGSCZJ1D13","proposition-management","Record claims, risks, beliefs, world facts, WSJF estimates, and assumptions through the Proposition interface.","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Fsrc\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fproposition-management\u002FSKILL.md",[],[251,211,252,253,42,254,255,256,38,29,257,258,218,259,219,260,261,32,262,37],"proposition management","proposition","interface","claims","record","world","creation","event","WSJF","index","updates","facts"]