[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6WZN5RMDR1SR67SB7YNP5FG":3,"skill-same:01M6WZN5RMDR1SR67SB7YNP5FG":102},{"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":39,"linkOnly":40,"markdownContent":41,"markdownBody":42,"frontMatter":43,"refName":69,"commitSha":70,"skillDirectory":71,"forksCount":14,"repositoryHtmlUrl":72,"repositoryHomepageUrl":73,"ownerHtmlUrl":74,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":75,"install":76,"resources":89},"01M6WZN5RMDR1SR67SB7YNP5FG","context-grooming",null,"Review and prune the project context — archive completed work, summarize stale entries, and propose disabling unused skills.","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\u002Farchive\u002Fprocesskit-install-20261101\u002Fcontext\u002Ftemplates\u002Fprocesskit\u002Fv0.18.1\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fcontext-grooming\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],"context grooming","context","entries","audit trail","unused","Existing","v1","event","keep","fresh","project","review","compact","query","log","work","archive","sqlite","audit","run","2026-11-10 10:12:30",false,"---\nname: context-grooming\ndescription: >\n  Review and prune the project context — archive completed work,\n  summarize stale entries, and propose disabling unused skills.\nmetadata:\n  processkit:\n    apiVersion: processkit.projectious.work\u002Fv1\n    id: SKILL-context-grooming\n    version: \"1.0.0\"\n    created: 2026-04-07T00:00:00Z\n    category: processkit\n    layer: 4\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    provides:\n      primitives: []\n      templates: [grooming-report]\n    commands:\n      - name: context-grooming-run\n        args: \"\"\n        description: \"Run a context grooming pass to prune and compact project context\"\n---\n\n# Context Grooming\n\n## Intro\n\n`context-grooming` is the periodic-cleanup skill. Once a week or once a\nmonth, the agent walks the project's `context\u002F` directory against a set\nof rules and produces a report listing what could be archived, summarized,\nor disabled. The user approves each item before any file moves. This is\nhow the project keeps its session context lean over time.\n\n## Overview\n\n### When to run\n\nTriggered by:\n- An explicit user request: \"groom the context\", \"let's clean up\"\n- A cadence configured in `processkit.toml`:\n  ```toml\n  [context.grooming]\n  cadence = \"weekly\"     # weekly | monthly | quarterly | never\n  ```\n  The agent checks the cadence at session start. If the last grooming was\n  more than the cadence ago, it surfaces:\n  > \"Context grooming was last run 5 weeks ago. Want me to do a pass?\"\n- A noticeable session-context bloat: more than ~16k tokens loaded at\n  session start (if the agent can detect this — depends on the runtime).\n\n### What grooming does\n\nGrooming is a **review-then-act** workflow:\n\n1. **Walk** — agent walks `context\u002F` against the ruleset (see Level 3).\n2. **Report** — agent produces a `grooming-report.md` document listing\n   proposed actions, organized by category.\n3. **Review** — user reads the report and approves items individually\n   (or in groups).\n4. **Act** — agent moves\u002Farchives\u002Fdisables only the approved items.\n5. **Record** — appends a `LogEntry` (`event_type: context.groomed`) and\n   updates the report's status field.\n\n**Nothing is moved silently. Nothing is deleted, ever** — the worst that\ncan happen is \"moved to `context\u002Farchive\u002F`\" which is reversible.\n\nThis skill also provides the `\u002Fcontext-grooming-run` slash command for direct invocation — see `commands\u002Fcontext-grooming-run.md`.\n\n### The grooming report\n\nThe report lives at `context\u002Fgrooming-reports\u002F\u003Cdate>.md`. Sample shape:\n\n```markdown\n---\napiVersion: processkit.projectious.work\u002Fv1\nkind: Context\nmetadata:\n  id: GROOM-2026-04-07\n  privacy: project-private\nspec:\n  state: pending  # pending | approved | applied\n  generated: 2026-04-07T10:00:00Z\n---\n\n# Grooming report — 2026-04-07\n\n## Workitems eligible for archive (12)\n\n| ID | Title | State | Last touched |\n|---|---|---|---|\n| BACK-foo | ... | done | 2026-02-01 |\n| ... | ... | ... | ... |\n\n[ ] Approve all\n[ ] Approve individually: ...\n\n## Standups eligible for summarization (8)\n\n8 standup entries from Q1 2026 can be summarized into a single quarterly\nsummary:\n- 2026-01-05.md → 2026-03-30.md\n\n[ ] Approve summarization\n\n## Unused skills (3)\n\nThese skills have not been invoked in any session for 60+ days:\n- `pixijs-gamedev` (last used: 2026-01-12)\n- `flutter-development` (last used: never since project init)\n- `latex-authoring` (last used: 2026-02-08)\n\n[ ] Approve disabling (move from context\u002Fskills\u002F to context\u002Fskills\u002F.disabled\u002F)\n```\n\n### The default ruleset\n\n| Rule | Action proposed |\n|---|---|\n| WorkItem in terminal state (`done` \u002F `cancelled`) for >30 days | Move to `context\u002Farchive\u002Fworkitems\u002F` |\n| LogEntry older than 90 days, not referenced by any current entity | Move to `context\u002Farchive\u002Flogs\u002F` |\n| Standup entry older than current quarter | Summarize all entries from a quarter into one quarterly summary; archive originals |\n| DecisionRecord superseded for >180 days | Move both the old and the new together to `context\u002Farchive\u002Fdecisions\u002F` |\n| File over 500 lines that hasn't been edited in 30 days | Propose splitting or summarizing |\n| Skill in `context\u002Fskills\u002F` not invoked in any session for 60 days | Propose disabling (move to `context\u002Fskills\u002F.disabled\u002F`) |\n| Migration in `applied\u002F` older than 1 year | Archive (rare — kept for historical reference) |\n\nThese are defaults. Projects can override in `processkit.toml`:\n\n```toml\n[context.grooming.rules]\nworkitem_archive_age_days = 60       # default 30\nlog_archive_age_days = 180           # default 90\nunused_skill_age_days = 90           # default 60\ndisable_skills_automatically = false # default false (always proposes, never auto-disables)\n```\n\n## Gotchas\n\nAgent-specific failure modes — provider-neutral pause-and-self-check items:\n\n- **Moving or deleting files without user approval.** Grooming is a propose-then-act workflow: generate the report, surface it to the user, act only on approved items. Silently moving files based on the ruleset — even files that clearly qualify — bypasses the user's judgment. Nothing moves without approval.\n- **Proposing to disable a skill that is currently being used.** The \"not invoked in 60 days\" rule may fire on a skill that was used extensively until recently, or that is central to a project the user is about to resume. Before proposing to disable a skill, check recent session notes and the backlog for references to that skill.\n- **Grooming the current quarter's standups or in-progress work.** The ruleset explicitly skips anything in-progress or less than 30 days old. Proposing to archive a standup from last week, or a workitem currently being worked on, produces a report that damages trust in the grooming process. Always apply the age and status filters strictly.\n- **Summarizing standups without preserving specific work item references.** A quarterly summary that replaces eight individual standup entries must retain all workitem IDs, decision references, and blocker resolutions mentioned in the originals. A summary that says \"worked on various items\" loses the audit trail. Summarizations are compressions, not erasures.\n- **Running grooming too frequently, making every report feel like noise.** If the user is asked to approve a grooming report every session, they will start ignoring or dismissing them reflexively. Respect the configured cadence. If the user has declined grooming three sessions in a row, offer to change the cadence rather than repeating the offer.\n- **Treating the default ruleset as mandatory even when the project has overrides.** The `processkit.toml` configuration can change thresholds — a project may archive workitems after 60 days instead of 30, or may never disable skills. Always read the configuration before applying any rule, and apply the project-specific values.\n- **Not recording the grooming event in the event log.** Every grooming session that results in file moves must be recorded as a `LogEntry` with `event_type: context.groomed`, listing which items were moved and when. Without this, the next grooming pass cannot determine when the last grooming ran.\n\n## Full reference\n\n### Configuration\n\n```toml\n[context.grooming]\ncadence = \"monthly\"           # weekly | monthly | quarterly | never (default: never)\nreport_dir = \"context\u002Fgrooming-reports\"   # default\n\n[context.grooming.rules]\n# Override individual rules. All have safe defaults.\nworkitem_archive_age_days = 30\nlogentry_archive_age_days = 90\nstandup_archive_age_days = 90\ndecision_archive_age_days = 180\noversize_file_threshold_lines = 500\nunused_skill_age_days = 60\n```\n\n### Archiving entities vs. archiving files\n\nIn the entity-sharded world, \"archiving\" an entity means **transitioning\nit to a terminal state** in its state machine — `done`, `cancelled`, or\n`archived` depending on the kind. This is the primary archive operation:\n\n- WorkItem → `done` or `cancelled`\n- DecisionRecord → `superseded` or `archived`\n- Note → `archived` or `promoted`\n\nOnce in a terminal state, the entity is filtered out of active queries\nby the MCP server (`query_entities(state=active)` returns nothing in\nterminal states). The file stays in place — no move needed for functional\npurposes.\n\n**Physical compaction** (moving terminal-state files to\n`context\u002F\u003Ckind>\u002Farchive\u002F`) is an optional aesthetic operation that\ncontext-grooming can propose when directories grow unwieldy for human\nbrowsing. The threshold is configured via `oversize_file_threshold_lines`\nor a simple entity count. Propose compaction — never act without approval.\n\n### What grooming will NOT touch\n\nTo stay safe and predictable, grooming explicitly skips:\n\n- Anything in `context\u002Fowner\u002F` (the personal context portfolio)\n- Anything in any `private\u002F` directory (user-private content)\n- The current quarter's standups (only previous quarters get summarized)\n- Currently in-progress work (Migration in `in-progress\u002F`, WorkItem in\n  `in-progress` state, etc.)\n- Files less than 30 days old (regardless of size)\n- The `context\u002Fmigrations\u002F` directory (migrations have their own lifecycle)\n\n### When to override the cadence\n\nSome projects benefit from rare grooming (long-running R&D where\nhistorical context matters); some benefit from frequent grooming (fast-\nmoving sprints). The cadence config is per-project.\n\nIf the user repeatedly says \"skip grooming\" three sessions in a row, the\nagent should propose:\n> \"You've skipped grooming three times. Do you want me to switch the\n> cadence to a longer interval (or `never`)?\"\n\n### Relationship to context budget (in `agent-management`)\n\nGrooming reduces the LIVE size of the context. The `[context.budget]`\nsection in `processkit.toml` (documented in `agent-management\u002FSKILL.md`) tells\nthe agent how much context to load per session. The two work together:\ngrooming makes the corpus smaller; budget tells the agent how to load\nselectively.\n\n### Ideas for future versions\n\n- **Ranking**: instead of binary \"archive or not\", rank files by\n  \"loading-cost vs information-value\" and propose the bottom 10%.\n- **Compaction**: instead of archive, the agent could rewrite a 500-line\n  workitem-with-history into a 20-line summary + a link to the archived\n  full version.\n- **Auto-grooming for reversible moves only**: a low-risk subset of the\n  rules could be applied automatically (still logged for review). v1\n  ships with always-propose, never-auto.\n","# Context Grooming\n\n## Intro\n\n`context-grooming` is the periodic-cleanup skill. Once a week or once a\nmonth, the agent walks the project's `context\u002F` directory against a set\nof rules and produces a report listing what could be archived, summarized,\nor disabled. The user approves each item before any file moves. This is\nhow the project keeps its session context lean over time.\n\n## Overview\n\n### When to run\n\nTriggered by:\n- An explicit user request: \"groom the context\", \"let's clean up\"\n- A cadence configured in `processkit.toml`:\n  ```toml\n  [context.grooming]\n  cadence = \"weekly\"     # weekly | monthly | quarterly | never\n  ```\n  The agent checks the cadence at session start. If the last grooming was\n  more than the cadence ago, it surfaces:\n  > \"Context grooming was last run 5 weeks ago. Want me to do a pass?\"\n- A noticeable session-context bloat: more than ~16k tokens loaded at\n  session start (if the agent can detect this — depends on the runtime).\n\n### What grooming does\n\nGrooming is a **review-then-act** workflow:\n\n1. **Walk** — agent walks `context\u002F` against the ruleset (see Level 3).\n2. **Report** — agent produces a `grooming-report.md` document listing\n   proposed actions, organized by category.\n3. **Review** — user reads the report and approves items individually\n   (or in groups).\n4. **Act** — agent moves\u002Farchives\u002Fdisables only the approved items.\n5. **Record** — appends a `LogEntry` (`event_type: context.groomed`) and\n   updates the report's status field.\n\n**Nothing is moved silently. Nothing is deleted, ever** — the worst that\ncan happen is \"moved to `context\u002Farchive\u002F`\" which is reversible.\n\nThis skill also provides the `\u002Fcontext-grooming-run` slash command for direct invocation — see `commands\u002Fcontext-grooming-run.md`.\n\n### The grooming report\n\nThe report lives at `context\u002Fgrooming-reports\u002F\u003Cdate>.md`. Sample shape:\n\n```markdown\n---\napiVersion: processkit.projectious.work\u002Fv1\nkind: Context\nmetadata:\n  id: GROOM-2026-04-07\n  privacy: project-private\nspec:\n  state: pending  # pending | approved | applied\n  generated: 2026-04-07T10:00:00Z\n---\n\n# Grooming report — 2026-04-07\n\n## Workitems eligible for archive (12)\n\n| ID | Title | State | Last touched |\n|---|---|---|---|\n| BACK-foo | ... | done | 2026-02-01 |\n| ... | ... | ... | ... |\n\n[ ] Approve all\n[ ] Approve individually: ...\n\n## Standups eligible for summarization (8)\n\n8 standup entries from Q1 2026 can be summarized into a single quarterly\nsummary:\n- 2026-01-05.md → 2026-03-30.md\n\n[ ] Approve summarization\n\n## Unused skills (3)\n\nThese skills have not been invoked in any session for 60+ days:\n- `pixijs-gamedev` (last used: 2026-01-12)\n- `flutter-development` (last used: never since project init)\n- `latex-authoring` (last used: 2026-02-08)\n\n[ ] Approve disabling (move from context\u002Fskills\u002F to context\u002Fskills\u002F.disabled\u002F)\n```\n\n### The default ruleset\n\n| Rule | Action proposed |\n|---|---|\n| WorkItem in terminal state (`done` \u002F `cancelled`) for >30 days | Move to `context\u002Farchive\u002Fworkitems\u002F` |\n| LogEntry older than 90 days, not referenced by any current entity | Move to `context\u002Farchive\u002Flogs\u002F` |\n| Standup entry older than current quarter | Summarize all entries from a quarter into one quarterly summary; archive originals |\n| DecisionRecord superseded for >180 days | Move both the old and the new together to `context\u002Farchive\u002Fdecisions\u002F` |\n| File over 500 lines that hasn't been edited in 30 days | Propose splitting or summarizing |\n| Skill in `context\u002Fskills\u002F` not invoked in any session for 60 days | Propose disabling (move to `context\u002Fskills\u002F.disabled\u002F`) |\n| Migration in `applied\u002F` older than 1 year | Archive (rare — kept for historical reference) |\n\nThese are defaults. Projects can override in `processkit.toml`:\n\n```toml\n[context.grooming.rules]\nworkitem_archive_age_days = 60       # default 30\nlog_archive_age_days = 180           # default 90\nunused_skill_age_days = 90           # default 60\ndisable_skills_automatically = false # default false (always proposes, never auto-disables)\n```\n\n## Gotchas\n\nAgent-specific failure modes — provider-neutral pause-and-self-check items:\n\n- **Moving or deleting files without user approval.** Grooming is a propose-then-act workflow: generate the report, surface it to the user, act only on approved items. Silently moving files based on the ruleset — even files that clearly qualify — bypasses the user's judgment. Nothing moves without approval.\n- **Proposing to disable a skill that is currently being used.** The \"not invoked in 60 days\" rule may fire on a skill that was used extensively until recently, or that is central to a project the user is about to resume. Before proposing to disable a skill, check recent session notes and the backlog for references to that skill.\n- **Grooming the current quarter's standups or in-progress work.** The ruleset explicitly skips anything in-progress or less than 30 days old. Proposing to archive a standup from last week, or a workitem currently being worked on, produces a report that damages trust in the grooming process. Always apply the age and status filters strictly.\n- **Summarizing standups without preserving specific work item references.** A quarterly summary that replaces eight individual standup entries must retain all workitem IDs, decision references, and blocker resolutions mentioned in the originals. A summary that says \"worked on various items\" loses the audit trail. Summarizations are compressions, not erasures.\n- **Running grooming too frequently, making every report feel like noise.** If the user is asked to approve a grooming report every session, they will start ignoring or dismissing them reflexively. Respect the configured cadence. If the user has declined grooming three sessions in a row, offer to change the cadence rather than repeating the offer.\n- **Treating the default ruleset as mandatory even when the project has overrides.** The `processkit.toml` configuration can change thresholds — a project may archive workitems after 60 days instead of 30, or may never disable skills. Always read the configuration before applying any rule, and apply the project-specific values.\n- **Not recording the grooming event in the event log.** Every grooming session that results in file moves must be recorded as a `LogEntry` with `event_type: context.groomed`, listing which items were moved and when. Without this, the next grooming pass cannot determine when the last grooming ran.\n\n## Full reference\n\n### Configuration\n\n```toml\n[context.grooming]\ncadence = \"monthly\"           # weekly | monthly | quarterly | never (default: never)\nreport_dir = \"context\u002Fgrooming-reports\"   # default\n\n[context.grooming.rules]\n# Override individual rules. All have safe defaults.\nworkitem_archive_age_days = 30\nlogentry_archive_age_days = 90\nstandup_archive_age_days = 90\ndecision_archive_age_days = 180\noversize_file_threshold_lines = 500\nunused_skill_age_days = 60\n```\n\n### Archiving entities vs. archiving files\n\nIn the entity-sharded world, \"archiving\" an entity means **transitioning\nit to a terminal state** in its state machine — `done`, `cancelled`, or\n`archived` depending on the kind. This is the primary archive operation:\n\n- WorkItem → `done` or `cancelled`\n- DecisionRecord → `superseded` or `archived`\n- Note → `archived` or `promoted`\n\nOnce in a terminal state, the entity is filtered out of active queries\nby the MCP server (`query_entities(state=active)` returns nothing in\nterminal states). The file stays in place — no move needed for functional\npurposes.\n\n**Physical compaction** (moving terminal-state files to\n`context\u002F\u003Ckind>\u002Farchive\u002F`) is an optional aesthetic operation that\ncontext-grooming can propose when directories grow unwieldy for human\nbrowsing. The threshold is configured via `oversize_file_threshold_lines`\nor a simple entity count. Propose compaction — never act without approval.\n\n### What grooming will NOT touch\n\nTo stay safe and predictable, grooming explicitly skips:\n\n- Anything in `context\u002Fowner\u002F` (the personal context portfolio)\n- Anything in any `private\u002F` directory (user-private content)\n- The current quarter's standups (only previous quarters get summarized)\n- Currently in-progress work (Migration in `in-progress\u002F`, WorkItem in\n  `in-progress` state, etc.)\n- Files less than 30 days old (regardless of size)\n- The `context\u002Fmigrations\u002F` directory (migrations have their own lifecycle)\n\n### When to override the cadence\n\nSome projects benefit from rare grooming (long-running R&D where\nhistorical context matters); some benefit from frequent grooming (fast-\nmoving sprints). The cadence config is per-project.\n\nIf the user repeatedly says \"skip grooming\" three sessions in a row, the\nagent should propose:\n> \"You've skipped grooming three times. Do you want me to switch the\n> cadence to a longer interval (or `never`)?\"\n\n### Relationship to context budget (in `agent-management`)\n\nGrooming reduces the LIVE size of the context. The `[context.budget]`\nsection in `processkit.toml` (documented in `agent-management\u002FSKILL.md`) tells\nthe agent how much context to load per session. The two work together:\ngrooming makes the corpus smaller; budget tells the agent how to load\nselectively.\n\n### Ideas for future versions\n\n- **Ranking**: instead of binary \"archive or not\", rank files by\n  \"loading-cost vs information-value\" and propose the bottom 10%.\n- **Compaction**: instead of archive, the agent could rewrite a 500-line\n  workitem-with-history into a 20-line summary + a link to the archived\n  full version.\n- **Auto-grooming for reversible moves only**: a low-risk subset of the\n  rules could be applied automatically (still logged for review). v1\n  ships with always-propose, never-auto.\n",{"name":5,"metadata":44,"description":68},{"processkit":45},{"id":46,"uses":47,"layer":54,"created":55,"version":56,"category":57,"commands":58,"provides":63,"apiVersion":67},"SKILL-context-grooming",[48,51],{"skill":49,"purpose":50},"event-log","Log events to keep the audit trail accurate after every write.",{"skill":52,"purpose":53},"index-management","Query existing entities and keep the SQLite index fresh after writes.",4,1775520000000,"1.0.0","processkit",[59],{"args":60,"name":61,"description":62},"","context-grooming-run","Run a context grooming pass to prune and compact project context",{"templates":64,"primitives":66},[65],"grooming-report",[],"processkit.projectious.work\u002Fv1","Review and prune the project context — archive completed work, summarize stale entries, and propose disabling unused skills.\n","main","4876a92620ee792aac2ac9a1a806322d8a572367","archive\u002Fprocesskit-install-20261101\u002Fcontext\u002Ftemplates\u002Fprocesskit\u002Fv0.18.1\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fcontext-grooming","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":77,"prompt":87,"zipAvailable":88},[78,81,84],{"tool":79,"command":80},"npx","npx skills add projectious-work\u002Fprocesskit --skill context-grooming",{"tool":82,"command":83},"bunx","bunx skills add projectious-work\u002Fprocesskit --skill context-grooming",{"tool":85,"command":86},"pnpm","pnpm dlx skills add projectious-work\u002Fprocesskit --skill context-grooming","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Farchive\u002Fprocesskit-install-20261101\u002Fcontext\u002Ftemplates\u002Fprocesskit\u002Fv0.18.1\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fcontext-grooming\u002FSKILL.md",true,[90,96],{"path":91,"resourceType":92,"resourceTypeDesc":93,"sizeBytes":94,"sourceHtmlUrl":95},"assets\u002Fgrooming-report.md","ASSET","资产文件",2129,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Farchive\u002Fprocesskit-install-20261101\u002Fcontext\u002Ftemplates\u002Fprocesskit\u002Fv0.18.1\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fcontext-grooming\u002Fassets\u002Fgrooming-report.md",{"path":97,"resourceType":98,"resourceTypeDesc":99,"sizeBytes":100,"sourceHtmlUrl":101},"commands\u002Fpk-groom.md","REFERENCE","参考文档",126,"https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002Farchive\u002Fprocesskit-install-20261101\u002Fcontext\u002Ftemplates\u002Fprocesskit\u002Fv0.18.1\u002Fcontext\u002Fskills\u002Fprocesskit\u002Fcontext-grooming\u002Fcommands\u002Fpk-groom.md",[103,110,117,123,149,169],{"skillNo":104,"name":105,"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":106,"categories":107,"keywords":108,"publishedAt":109},"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":111,"name":112,"nameZh":6,"description":113,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":114,"categories":115,"keywords":116,"publishedAt":109},"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":118,"name":119,"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":120,"categories":121,"keywords":122,"publishedAt":109},"01M76650NN4T8P241021NBR64S","standard-agent-building","https:\u002F\u002Fgithub.com\u002Fprojectious-work\u002Fprocesskit\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fstandard-agent-building\u002FSKILL.md",[],[],{"skillNo":124,"name":125,"nameZh":6,"description":126,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":127,"categories":128,"keywords":134,"publishedAt":39},"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",[129],{"categoryNo":130,"categoryName":131,"taxonomyType":132,"taxonomyTypeDesc":133},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[135,136,137,25,138,139,34,140,141,142,143,144,125,145,146,147,148],"management","skill management","ontology","create","as","transition","get","manage","update","alpha","list","package","skill package","manifests",{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":156,"publishedAt":39},"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",[155],{"categoryNo":130,"categoryName":131,"taxonomyType":132,"taxonomyTypeDesc":133},[157,158,135,159,160,161,162,163,164,165,166,167,34,141,142,144,168],"schema","schema management","contract","rebuild","contracts","workflows","wed","generated","MCP","validation","deterministic","mode",{"skillNo":170,"name":171,"nameZh":6,"description":172,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":173,"categories":174,"keywords":175,"publishedAt":39},"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",[],[176,135,177,178,138,179,180,181,34,141,182,26,143,183,144,184,185,32,186,33],"proposition management","proposition","interface","claims","record","world","creation","WSJF","index","updates","facts"]