[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DSTZP1YM56YY80J85NXT11":3,"skill-same:01M6DSTZP1YM56YY80J85NXT11":92},{"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":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":43,"commitSha":44,"skillDirectory":45,"forksCount":46,"repositoryHtmlUrl":47,"repositoryHomepageUrl":48,"ownerHtmlUrl":49,"ownerBlogUrl":50,"ownerDescription":6,"updatedAt":51,"install":52,"resources":65},"01M6DSTZP1YM56YY80J85NXT11","org-memory-synthesis",null,"Synthesize org-memory — fold new events into recent.md, promote cross-repo patterns to decisions.md and rules.md, and audit for staleness, cross-file conflicts, and drift. Invoke whenever the user asks \"synthesize org-memory\", \"fold events\", \"is recent.md stale\", \"audit org-memory\", or any phrasing like \"did we synthesize org-memory\" — even if they do not explicitly type the slash command.","kiloloop\u002Foacp-skills","kiloloop","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F264514437?v=4","01M6DSS1K2JKZD3YRMXRH54968","Kiloloop","01M6DSS1SMM810F7327W2YX7ZZ",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Forg-memory-synthesis\u002Fcodex\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"org memory synthesis","Synthesis","memory","stale","type","synthesize","drift","patterns","do","conflicts","Fold","user","command","decisions","repo","Rules","promote","audit","file","2026-11-10 10:12:30",false,"---\nname: org-memory-synthesis\ndescription: Synthesize org-memory — fold new events into recent.md, promote cross-repo patterns to decisions.md and rules.md, and audit for staleness, cross-file conflicts, and drift. Invoke whenever the user asks \"synthesize org-memory\", \"fold events\", \"is recent.md stale\", \"audit org-memory\", or any phrasing like \"did we synthesize org-memory\" — even if they do not explicitly type the slash command.\n---\n\n# \u002Forg-memory-synthesis — Org-Memory Synthesis\n\nOrg-memory (`$OACP_HOME\u002Forg-memory\u002F`) is the cross-project SSOT that the OACP\nsession-init hook auto-loads. It drifts if events accumulate unfolded and\ntopical files (`decisions.md`, `rules.md`) are not curated. This skill does the\nfull 3-file pass — synthesis then audit — so the curated layer stays\ntrustworthy.\n\n## When to use\n\nUse this skill when:\n\n- The user asks whether org-memory is current\n- A wrap-up or housekeeping flow needs to fold recent events\n- Cross-file inconsistency or stale entries surface during another task\n- An upstream skill hands off the org-memory pass\n\n## Files & paths\n\n| Path | Role |\n|------|------|\n| `$OACP_HOME\u002Forg-memory\u002Frecent.md` | Rolling current-state, target ≤150 lines, auto-loaded by session-init hook |\n| `$OACP_HOME\u002Forg-memory\u002Fdecisions.md` | Architectural decisions (cross-repo, 3+ pattern threshold) |\n| `$OACP_HOME\u002Forg-memory\u002Frules.md` | Standing conventions (cross-repo, 3+ pattern threshold) |\n| `$OACP_HOME\u002Forg-memory\u002Fevents\u002FYYYYMMDD-HHMMSS-\u003Cslug>.md` | Raw event files with YAML frontmatter (date, agent, project, type, related) |\n\n`recent.md` header carries `\u003C!-- Last event folded in: \u003Cfilename> -->`. Use it\nto scope incremental synthesis cheaply. `OACP_HOME` defaults to `$HOME\u002Foacp`\nper the OACP CLI; check the workspace marker if unsure.\n\n## Setup\n\nResolve the workspace and helper before reading event bodies:\n\n```bash\nOACP_ROOT=\"${OACP_HOME:-$HOME\u002Foacp}\"\ntest -d \"$OACP_ROOT\u002Forg-memory\" || {\n  echo \"ERROR: org-memory not found. Run: agent-memory org init --home \\\"$OACP_ROOT\\\"\" >&2\n  exit 1\n}\n```\n\n## Synthesis flow\n\n### 1. Scope the work\n\nList event filenames newer than the marker in `recent.md`. Filenames are\nlex-sortable on the `YYYYMMDD-HHMMSS-` prefix, so an `awk '$0 > m'` filter is\nthe comparison:\n\n```bash\nOACP_ROOT=\"${OACP_HOME:-$HOME\u002Foacp}\"\nRECENT=\"$OACP_ROOT\u002Forg-memory\u002Frecent.md\"\nEVENTS_DIR=\"$OACP_ROOT\u002Forg-memory\u002Fevents\"\n\nMARKER=$(command grep -m1 \"Last event folded in:\" \"$RECENT\" \\\n  | sed -E 's\u002F.*Last event folded in:[[:space:]]+([^[:space:]]+).*\u002F\\1\u002F')\n\nif [ -z \"$MARKER\" ]; then\n  echo \"WARN: no marker in $RECENT header; listing ALL events (full pass)\" >&2\n  command ls \"$EVENTS_DIR\" | sort\nelse\n  command ls \"$EVENTS_DIR\" | sort | awk -v m=\"$MARKER\" '$0 > m'\nfi\n```\n\n- Empty output means org-memory is current; jump to Step 7 for the audit-only\n  pass.\n- Missing marker in older `recent.md` means the inline shell lists all events\n  and warns; do a full pass and tell the user the read cost.\n\n### 2. Batch-read new event bodies\n\nUse one shell read across the listed event files:\n\n```bash\ncd $OACP_HOME\u002Forg-memory\u002Fevents && for f in \u003Cnewline-separated names from step 1>; do echo \"=== $f ===\"; command tail -n +9 \"$f\"; echo; done\n```\n\n`tail -n +9` skips the YAML frontmatter. Use the frontmatter for routing; fold\nthe body into the curated files.\n\n### 3. Categorize each event\n\nThe event's `type:` frontmatter field drives placement:\n\n| `type` | Default placement in `recent.md` | Promotion candidate? |\n|--------|----------------------------------|----------------------|\n| `event` | Current State | Rarely — only milestones with durable cross-project bearing |\n| `rule` | Standing Rules | Promote to `rules.md` if cross-repo (3+ repos in `related:`) or generalizable convention |\n| `decision` | Active Decisions | Promote to `decisions.md` if architectural or cross-repo |\n\nCross-repo signal: count distinct repos in the event's `related:` field. 3+ is\nthe promotion threshold. Skill-specific rules that affect only one skill's flow\nstay in `recent.md` Standing Rules; promoting them adds `rules.md` noise without\ncross-repo benefit.\n\n### 4. Propose a 3-file diff to the user\n\nPresent the intended changes as three sections: `recent.md`, `decisions.md`,\nand `rules.md`. For each promotion, include a one-line rationale tied to the 3+\npattern threshold, cross-repo signal, or convention class. Note `recent.md`\nline-budget impact so it stays under 150 lines.\n\nIf all edits are mechanical and uncontroversial, proceed after the proposal.\nPause for judgment calls where local context cannot resolve the truth.\n\n### 5. Apply edits\n\nUse Codex's normal file-editing flow (`apply_patch` for manual edits) and keep\nthe change set limited to:\n\n- `$OACP_HOME\u002Forg-memory\u002Frecent.md`\n- `$OACP_HOME\u002Forg-memory\u002Fdecisions.md`\n- `$OACP_HOME\u002Forg-memory\u002Frules.md`\n\nPlace new `decisions.md` entries at the **top** of the file under the correct\n`## YYYY-MM-DD` heading so newest-first order is preserved. Step 7D will\nmechanically verify order.\n\n### 6. Update the markers\n\nTwo header lines in `recent.md` must stay present:\n\n- `\u003C!-- Last event folded in: \u003Cnewest event filename from step 1> -->`\n- `\u003C!-- Last curated: YYYY-MM-DD by \u003Cwho> -->`\n\nA dropped marker forces the next pass to re-scan all events, defeating the\nincremental design.\n\n### 7. Audit pass — ALWAYS run after synthesis\n\nSix checks, in order. The audit catches what synthesis alone misses.\n\n**A. Cross-file consistency.** For items appearing in both `recent.md` and a\ncurated file, compare key facts such as product structure, paths, versions, and\nstatus claims. Surface mismatches with file references.\n\n**B. Supersession check.** Scan `decisions.md` for `Supersedes X` annotations.\nFor each X, verify X is also marked as superseded with a pointer back.\nAsymmetric supersession reads as if both entries are active.\n\n**C. Stale-status check.** Find entries with status claims like \"Filed #N\",\n\"Pending review\", or \"Q1 target\". For GitHub issue refs where the repo can be\ninferred, verify state:\n\n```bash\ngh issue view \u003CN> --repo \u003Cowner>\u002F\u003Crepo> --json state,closedAt,updatedAt\n```\n\nAnnotate stale claims inline rather than removing them, for example:\n`Status: #15 still OPEN P3, not actioned`.\n\n**D. Chronological order in `decisions.md`.** Section headers must run\n**newest-first, top-to-bottom** so the freshest decisions are read first. Run\nthe mechanical check — do not self-report:\n\n```bash\nDECISION_HEADERS=\"$(mktemp)\"\ncommand grep -nE '^## 20' \"$OACP_HOME\u002Forg-memory\u002Fdecisions.md\" \\\n  | awk '{print $2}' >\"$DECISION_HEADERS\" || true\nif [ ! -s \"$DECISION_HEADERS\" ]; then\n  echo \"No dated decision headers found\" >&2\n  command rm -f -- \"$DECISION_HEADERS\"\n  exit 1\nfi\nORDER_RC=0\nLC_ALL=C sort -rc \"$DECISION_HEADERS\" || ORDER_RC=$?\ncommand rm -f -- \"$DECISION_HEADERS\"\ntest \"$ORDER_RC\" -eq 0\n```\n\nExit zero means descending order is correct. If `sort -rc` reports disorder or\nno dated headers exist, repair with one edit operation. The mechanical check is\nthe verification; never substitute self-reporting.\n\n**E. Category drift.** Conventions filed under `decisions.md` should move to\n`rules.md`, and architectural decisions filed under `rules.md` should move to\n`decisions.md`. If unclear, leave the item where the historical event-author\nput it and mention the ambiguity.\n\n**F. Migration leftovers.** Content duplicated with another memory tree, such\nas a domain-specific memory store, should be trimmed to a single pointer.\n\n### 8. Surface findings, apply, ask\n\nApply uncontroversial fixes directly: annotations, chronological re-sort, and\nstrict duplicate removal. Pause and ask on judgment calls such as\npricing\u002Fstatus conflicts or ownership ambiguity.\n\n### 9. Suggest commit; do not auto-commit\n\nTypical commit message:\n\n```text\norg-memory: synthesize \u003CN> events + audit pass\n\n- \u003CN> events folded into recent.md (\u003CM> to decisions.md, \u003CK> to rules.md)\n- \u003Caudit findings summary>\n```\n\nLet the user decide commit timing. The skill never commits automatically.\n\n## Anti-patterns\n\n- Do not delete superseded entries; annotate them. History is load-bearing.\n- Do not promote skill-specific rules to `rules.md`.\n- Do not auto-commit.\n- Do not edit sibling memory trees or project-scoped memory directories.\n- Do not drop the marker in `recent.md`.\n- Do not replace the Step 7D `sort -rc` check with a prose claim.\n\n## Workarounds\n\n- Use `command ls`, `command grep`, and `command tail` to bypass shell aliases.\n- `gh issue view` in Step 7C may need authenticated GitHub CLI access; if auth\n  is missing, report the unchecked refs instead of inventing status.\n- `recent.md` target is ≤150 lines; `decisions.md` and `rules.md` grow over\n  time. Trim by annotation or consolidation, not by deleting history.\n\n## Notes\n\n- The skill operates on `$OACP_HOME\u002Forg-memory\u002F` only.\n- Step 1's scope-the-work shell is inlined directly in this file (and in the\n  Claude top-level `SKILL.md`) — there is no separate helper script.\n- `shared\u002FINTENT.md` is the source of truth for runtime-agnostic behavior.\n\n## Learned from runs\n\n\u003C!-- Add post-run lessons here once \u002Forg-memory-synthesis has shipped and accumulated real-world feedback. -->\n","# \u002Forg-memory-synthesis — Org-Memory Synthesis\n\nOrg-memory (`$OACP_HOME\u002Forg-memory\u002F`) is the cross-project SSOT that the OACP\nsession-init hook auto-loads. It drifts if events accumulate unfolded and\ntopical files (`decisions.md`, `rules.md`) are not curated. This skill does the\nfull 3-file pass — synthesis then audit — so the curated layer stays\ntrustworthy.\n\n## When to use\n\nUse this skill when:\n\n- The user asks whether org-memory is current\n- A wrap-up or housekeeping flow needs to fold recent events\n- Cross-file inconsistency or stale entries surface during another task\n- An upstream skill hands off the org-memory pass\n\n## Files & paths\n\n| Path | Role |\n|------|------|\n| `$OACP_HOME\u002Forg-memory\u002Frecent.md` | Rolling current-state, target ≤150 lines, auto-loaded by session-init hook |\n| `$OACP_HOME\u002Forg-memory\u002Fdecisions.md` | Architectural decisions (cross-repo, 3+ pattern threshold) |\n| `$OACP_HOME\u002Forg-memory\u002Frules.md` | Standing conventions (cross-repo, 3+ pattern threshold) |\n| `$OACP_HOME\u002Forg-memory\u002Fevents\u002FYYYYMMDD-HHMMSS-\u003Cslug>.md` | Raw event files with YAML frontmatter (date, agent, project, type, related) |\n\n`recent.md` header carries `\u003C!-- Last event folded in: \u003Cfilename> -->`. Use it\nto scope incremental synthesis cheaply. `OACP_HOME` defaults to `$HOME\u002Foacp`\nper the OACP CLI; check the workspace marker if unsure.\n\n## Setup\n\nResolve the workspace and helper before reading event bodies:\n\n```bash\nOACP_ROOT=\"${OACP_HOME:-$HOME\u002Foacp}\"\ntest -d \"$OACP_ROOT\u002Forg-memory\" || {\n  echo \"ERROR: org-memory not found. Run: agent-memory org init --home \\\"$OACP_ROOT\\\"\" >&2\n  exit 1\n}\n```\n\n## Synthesis flow\n\n### 1. Scope the work\n\nList event filenames newer than the marker in `recent.md`. Filenames are\nlex-sortable on the `YYYYMMDD-HHMMSS-` prefix, so an `awk '$0 > m'` filter is\nthe comparison:\n\n```bash\nOACP_ROOT=\"${OACP_HOME:-$HOME\u002Foacp}\"\nRECENT=\"$OACP_ROOT\u002Forg-memory\u002Frecent.md\"\nEVENTS_DIR=\"$OACP_ROOT\u002Forg-memory\u002Fevents\"\n\nMARKER=$(command grep -m1 \"Last event folded in:\" \"$RECENT\" \\\n  | sed -E 's\u002F.*Last event folded in:[[:space:]]+([^[:space:]]+).*\u002F\\1\u002F')\n\nif [ -z \"$MARKER\" ]; then\n  echo \"WARN: no marker in $RECENT header; listing ALL events (full pass)\" >&2\n  command ls \"$EVENTS_DIR\" | sort\nelse\n  command ls \"$EVENTS_DIR\" | sort | awk -v m=\"$MARKER\" '$0 > m'\nfi\n```\n\n- Empty output means org-memory is current; jump to Step 7 for the audit-only\n  pass.\n- Missing marker in older `recent.md` means the inline shell lists all events\n  and warns; do a full pass and tell the user the read cost.\n\n### 2. Batch-read new event bodies\n\nUse one shell read across the listed event files:\n\n```bash\ncd $OACP_HOME\u002Forg-memory\u002Fevents && for f in \u003Cnewline-separated names from step 1>; do echo \"=== $f ===\"; command tail -n +9 \"$f\"; echo; done\n```\n\n`tail -n +9` skips the YAML frontmatter. Use the frontmatter for routing; fold\nthe body into the curated files.\n\n### 3. Categorize each event\n\nThe event's `type:` frontmatter field drives placement:\n\n| `type` | Default placement in `recent.md` | Promotion candidate? |\n|--------|----------------------------------|----------------------|\n| `event` | Current State | Rarely — only milestones with durable cross-project bearing |\n| `rule` | Standing Rules | Promote to `rules.md` if cross-repo (3+ repos in `related:`) or generalizable convention |\n| `decision` | Active Decisions | Promote to `decisions.md` if architectural or cross-repo |\n\nCross-repo signal: count distinct repos in the event's `related:` field. 3+ is\nthe promotion threshold. Skill-specific rules that affect only one skill's flow\nstay in `recent.md` Standing Rules; promoting them adds `rules.md` noise without\ncross-repo benefit.\n\n### 4. Propose a 3-file diff to the user\n\nPresent the intended changes as three sections: `recent.md`, `decisions.md`,\nand `rules.md`. For each promotion, include a one-line rationale tied to the 3+\npattern threshold, cross-repo signal, or convention class. Note `recent.md`\nline-budget impact so it stays under 150 lines.\n\nIf all edits are mechanical and uncontroversial, proceed after the proposal.\nPause for judgment calls where local context cannot resolve the truth.\n\n### 5. Apply edits\n\nUse Codex's normal file-editing flow (`apply_patch` for manual edits) and keep\nthe change set limited to:\n\n- `$OACP_HOME\u002Forg-memory\u002Frecent.md`\n- `$OACP_HOME\u002Forg-memory\u002Fdecisions.md`\n- `$OACP_HOME\u002Forg-memory\u002Frules.md`\n\nPlace new `decisions.md` entries at the **top** of the file under the correct\n`## YYYY-MM-DD` heading so newest-first order is preserved. Step 7D will\nmechanically verify order.\n\n### 6. Update the markers\n\nTwo header lines in `recent.md` must stay present:\n\n- `\u003C!-- Last event folded in: \u003Cnewest event filename from step 1> -->`\n- `\u003C!-- Last curated: YYYY-MM-DD by \u003Cwho> -->`\n\nA dropped marker forces the next pass to re-scan all events, defeating the\nincremental design.\n\n### 7. Audit pass — ALWAYS run after synthesis\n\nSix checks, in order. The audit catches what synthesis alone misses.\n\n**A. Cross-file consistency.** For items appearing in both `recent.md` and a\ncurated file, compare key facts such as product structure, paths, versions, and\nstatus claims. Surface mismatches with file references.\n\n**B. Supersession check.** Scan `decisions.md` for `Supersedes X` annotations.\nFor each X, verify X is also marked as superseded with a pointer back.\nAsymmetric supersession reads as if both entries are active.\n\n**C. Stale-status check.** Find entries with status claims like \"Filed #N\",\n\"Pending review\", or \"Q1 target\". For GitHub issue refs where the repo can be\ninferred, verify state:\n\n```bash\ngh issue view \u003CN> --repo \u003Cowner>\u002F\u003Crepo> --json state,closedAt,updatedAt\n```\n\nAnnotate stale claims inline rather than removing them, for example:\n`Status: #15 still OPEN P3, not actioned`.\n\n**D. Chronological order in `decisions.md`.** Section headers must run\n**newest-first, top-to-bottom** so the freshest decisions are read first. Run\nthe mechanical check — do not self-report:\n\n```bash\nDECISION_HEADERS=\"$(mktemp)\"\ncommand grep -nE '^## 20' \"$OACP_HOME\u002Forg-memory\u002Fdecisions.md\" \\\n  | awk '{print $2}' >\"$DECISION_HEADERS\" || true\nif [ ! -s \"$DECISION_HEADERS\" ]; then\n  echo \"No dated decision headers found\" >&2\n  command rm -f -- \"$DECISION_HEADERS\"\n  exit 1\nfi\nORDER_RC=0\nLC_ALL=C sort -rc \"$DECISION_HEADERS\" || ORDER_RC=$?\ncommand rm -f -- \"$DECISION_HEADERS\"\ntest \"$ORDER_RC\" -eq 0\n```\n\nExit zero means descending order is correct. If `sort -rc` reports disorder or\nno dated headers exist, repair with one edit operation. The mechanical check is\nthe verification; never substitute self-reporting.\n\n**E. Category drift.** Conventions filed under `decisions.md` should move to\n`rules.md`, and architectural decisions filed under `rules.md` should move to\n`decisions.md`. If unclear, leave the item where the historical event-author\nput it and mention the ambiguity.\n\n**F. Migration leftovers.** Content duplicated with another memory tree, such\nas a domain-specific memory store, should be trimmed to a single pointer.\n\n### 8. Surface findings, apply, ask\n\nApply uncontroversial fixes directly: annotations, chronological re-sort, and\nstrict duplicate removal. Pause and ask on judgment calls such as\npricing\u002Fstatus conflicts or ownership ambiguity.\n\n### 9. Suggest commit; do not auto-commit\n\nTypical commit message:\n\n```text\norg-memory: synthesize \u003CN> events + audit pass\n\n- \u003CN> events folded into recent.md (\u003CM> to decisions.md, \u003CK> to rules.md)\n- \u003Caudit findings summary>\n```\n\nLet the user decide commit timing. The skill never commits automatically.\n\n## Anti-patterns\n\n- Do not delete superseded entries; annotate them. History is load-bearing.\n- Do not promote skill-specific rules to `rules.md`.\n- Do not auto-commit.\n- Do not edit sibling memory trees or project-scoped memory directories.\n- Do not drop the marker in `recent.md`.\n- Do not replace the Step 7D `sort -rc` check with a prose claim.\n\n## Workarounds\n\n- Use `command ls`, `command grep`, and `command tail` to bypass shell aliases.\n- `gh issue view` in Step 7C may need authenticated GitHub CLI access; if auth\n  is missing, report the unchecked refs instead of inventing status.\n- `recent.md` target is ≤150 lines; `decisions.md` and `rules.md` grow over\n  time. Trim by annotation or consolidation, not by deleting history.\n\n## Notes\n\n- The skill operates on `$OACP_HOME\u002Forg-memory\u002F` only.\n- Step 1's scope-the-work shell is inlined directly in this file (and in the\n  Claude top-level `SKILL.md`) — there is no separate helper script.\n- `shared\u002FINTENT.md` is the source of truth for runtime-agnostic behavior.\n\n## Learned from runs\n\n\u003C!-- Add post-run lessons here once \u002Forg-memory-synthesis has shipped and accumulated real-world feedback. -->\n",{"name":5,"description":7},"main","ea78d5cce0208a15b59f6f2c49e729ba71a3c5af","skills\u002Forg-memory-synthesis\u002Fcodex",1,"https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills","https:\u002F\u002Foacp.dev\u002Fskills\u002F","https:\u002F\u002Fgithub.com\u002Fkiloloop","https:\u002F\u002Fkiloloop.com","2026-10-31 14:39:52",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add kiloloop\u002Foacp-skills --skill org-memory-synthesis",{"tool":58,"command":59},"bunx","bunx skills add kiloloop\u002Foacp-skills --skill org-memory-synthesis",{"tool":61,"command":62},"pnpm","pnpm dlx skills add kiloloop\u002Foacp-skills --skill org-memory-synthesis","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Forg-memory-synthesis\u002Fcodex\u002FSKILL.md",true,[66,72,76,80,84,88],{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"evals\u002Fevals.json","REFERENCE","参考文档",7983,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Fevals\u002Fevals.json",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Floop-catalog.md",45458,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Freferences\u002Floop-catalog.md",{"path":77,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Floop-guardrails.md",4998,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Freferences\u002Floop-guardrails.md",{"path":81,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Floop-orchestration.md",4813,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Freferences\u002Floop-orchestration.md",{"path":85,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":86,"sourceHtmlUrl":87},"references\u002Floop-state.md",3818,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Freferences\u002Floop-state.md",{"path":89,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":90,"sourceHtmlUrl":91},"references\u002Floop-template.md",5122,"https:\u002F\u002Fgithub.com\u002Fpatinaproject\u002Fskills\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmarketing-loops\u002Freferences\u002Floop-template.md",[93,104,109,134,140,148],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":38},"01M6DSWR7WDRG7C3D31R4QKW5T","\u003Cskill-name>","\u003Cone-line description>","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Ftemplate\u002Fcodex\u002FSKILL.md",[],[100,101,102,103],"skill name","name","\u003Cskill name>","line",{"skillNo":105,"name":95,"nameZh":6,"description":96,"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":38},"01M6DSWMRPCHQH4HJJF3MATTXZ","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Ftemplate\u002Fclaude\u002FSKILL.md",[],[100,101,102,103],{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":38},"01M6DSWH83XKNGJSKX2XWQCRKJ","wrap-up","End-of-session cleanup for Codex — inspect stale artifacts, run debrief, write significant org-memory events, run a delta-first self-improve pass, then publish bounded approved memory and repository changes.","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwrap-up\u002Fcodex\u002FSKILL.md",[],[116,117,118,119,21,120,121,122,123,124,125,126,127,22,128,129,130,131,132,133],"wrap up","up","wrap","run","cleanup","artifacts","repository","publish","first","improve","write","session","session cleanup","delta","self","end","debrief","codex",{"skillNo":135,"name":136,"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":137,"categories":138,"keywords":139,"publishedAt":38},"01M6DSWEQBS2CE6F2H8T3VQNZY","claude","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwrap-up\u002Fclaude\u002FSKILL.md",[],[],{"skillNo":141,"name":111,"nameZh":6,"description":142,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":143,"categories":144,"keywords":145,"publishedAt":38},"01M6DSW99EPXDVAG8GHR785CC5","End-of-session cleanup, debrief, org-memory events, self-improve, commit, and push in one command.","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Fwrap-up\u002FSKILL.md",[],[116,117,118,130,21,146,120,147,125,31,127,128,131,132],"commit","push",{"skillNo":149,"name":150,"nameZh":6,"description":151,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":152,"categories":153,"keywords":154,"publishedAt":38},"01M6DSW4FRNRKRQ0TRMZAYNMKZ","self-improve","Audit and improve Codex operating guidance: skills, curated project memory, AGENTS.md, and relevant configuration. Use for explicit self-improve requests, routine wrap-up drift checks, stale or conflicting guidance, or an approved cleanup. Route release-wide OACP migrations and context-budget audits to audit-oacp-skills. Do not hand-edit generated Codex memories or installed plugin\u002Fsystem artifacts.","https:\u002F\u002Fgithub.com\u002Fkiloloop\u002Foacp-skills\u002Fblob\u002Fmain\u002Fskills\u002Fself-improve\u002Fcodex\u002FSKILL.md",[],[155,130,125,156,120,157,36,158,159,160,161,162,117,27,163,25,164,165,166,21],"self improve","project","release","Budget","system","edit","context","generated","route","memories","operating","oacp"]