[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71JSHE4AXK3PFM5HT7STR97":3,"skill-same:01M71JSHE4AXK3PFM5HT7STR97":158},{"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":28,"publishedAt":49,"linkOnly":50,"markdownContent":51,"markdownBody":52,"frontMatter":53,"refName":56,"commitSha":57,"skillDirectory":58,"forksCount":59,"repositoryHtmlUrl":60,"repositoryHomepageUrl":61,"ownerHtmlUrl":62,"ownerBlogUrl":63,"ownerDescription":64,"updatedAt":65,"install":66,"resources":79},"01M71JSHE4AXK3PFM5HT7STR97","arch-index",null,"Build and validate .arch\u002Findex.json — a committed map from each architectural component of this repository to the real directories and files that implement it, pinned to a git commit, with every path verified to exist. Also detects structural drift between that index and the current code. Use when asked to map a codebase, produce a component or code-ownership index, find which files implement a component, check whether architecture docs are stale, or detect architecture drift after a refactor. Triggers include 'map the codebase', 'component index', 'which files implement X', 'is the architecture still accurate', 'architecture drift', 'refresh the architecture index'.","coco-research\u002Fcoco","coco-research","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F68911604?v=4","01M71JRE9C4NM4NSZM4JD0BJTB","Coco Research","01M71JREA0ZMR0PMH3CKA99SN1",630,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002FSKILL.md",[18,23],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48],"arch index","index","arch","triggers","path","refactor","verified","git","repository","component","architecture","commit","code","drift","when","files","map","structural","directories","which","2026-11-10 10:12:30",false,"---\nname: arch-index\ndescription: \"Build and validate .arch\u002Findex.json — a committed map from each architectural component of this repository to the real directories and files that implement it, pinned to a git commit, with every path verified to exist. Also detects structural drift between that index and the current code. Use when asked to map a codebase, produce a component or code-ownership index, find which files implement a component, check whether architecture docs are stale, or detect architecture drift after a refactor. Triggers include 'map the codebase', 'component index', 'which files implement X', 'is the architecture still accurate', 'architecture drift', 'refresh the architecture index'.\"\ndomain: engineering\nversion: 1.0.0\n---\n\n# arch-index\n\nProduces one artifact: `.arch\u002Findex.json`, a commit-pinned map from each architectural\ncomponent of a repository to the real directories and files that implement it. Every\nclaimed path is verified to exist by a script with an exit code, not by a promise in a\nprompt.\n\n**Guiding principle — Paths or it does not exist.** The protocol that consumers follow\nis `team:architecture.md`. This file is the producer.\n\n## What this skill is for, and what it is not for\n\nIt is for answering, mechanically: which files implement this component, and has that\nstopped being true.\n\nIt is not a design tool. The index is **reverse-only** — it describes code that exists\non disk at a known commit. There is no mode for specifying a system that has not been\nbuilt, because the central check is that every path resolves. Forward design work\nbelongs in `\u002Fteam think`, the `SI-*-Design` councils, or\n`\u002Futil:create-architecture-documentation`.\n\nIt is not a codebase-mapping tool either, and it should not act like one. When\n`.planning\u002Fcodebase\u002F` exists, `gsd-map-codebase` has already produced seven structured\ndocuments covering the stack, integrations, architecture, structure, conventions,\ntesting, and concerns. This skill **reads those and skips crawling entirely.** Crawling\na repository that has already been mapped is the exact waste this skill was designed to\navoid, and duplicating `gsd-map-codebase` would make it a worse version of a tool that\nalready exists.\n\nIt is not a work-allocation mechanism. The index sits at C4-Container altitude, so a\nsingle component can own thousands of files. Nothing derives per-agent file ownership\nfrom it.\n\n## Modes\n\nResolve the mode from the invocation. When none is given, use `drift` if\n`.arch\u002Fpinned-commit` exists and `build` if it does not.\n\n### `build`\n\n**Step 1 — find prior art before doing any work.** In this order:\n\n1. If `.planning\u002Fcodebase\u002F` exists, read `ARCHITECTURE.md`, `STRUCTURE.md`,\n   `STACK.md`, and `INTEGRATIONS.md` from it. Set `source` to `gsd-codebase-map`. Skip\n   to step 2. Do not crawl.\n2. Otherwise, if gitnexus is available for this repository, call\n   `mcp__gitnexus__context` and `mcp__gitnexus__group_list` for a deterministic view of\n   module grouping. Set `source` to `gitnexus`. Skip to step 2.\n\n   A gitnexus index that is behind HEAD is still usable here, and `list_repos` reports\n   exactly how far behind it is. Record that number in `ecosystem.sourceCommitsBehind`.\n   The reason staleness is tolerable is structural rather than optimistic: gitnexus\n   informs *component grouping*, while every path the index claims is checked by the\n   validator against the working tree. A path that moved since gitnexus was indexed\n   fails check 5 and blocks the build. Stale grouping advice can therefore make the\n   decomposition slightly dated, but it cannot introduce a path that does not exist.\n   Still run `scripts\u002Frepo_tree.py` for the authoritative current tree.\n3. Otherwise crawl. Run `scripts\u002Frepo_tree.py --repo-root .` for a filtered,\n   depth-adaptive tree, then perform one exploration pass under\n   `references\u002Ftool-frugality.md`, writing `.arch\u002FANALYSIS.md` against the seven-section\n   skeleton in `references\u002Fanalysis-taxonomy.md` under its ten-thousand-character cap.\n   Set `source` to `crawl`.\n\nIf the tree came back truncated, say so at the top of the analysis and set\n`ecosystem.treeTruncated` to `true` in the index. A truncated tree disables `drift`\nmode entirely, by design.\n\n**Step 2 — synthesise, with no tool use.** Read only the gathered summary plus the\ntree. Emit `.arch\u002Findex.json` per `references\u002Fschema.md`, bound by\n`references\u002Fcomponent-rules.md`. This stage does not explore; separating exploration\nfrom formatting is why the structured output stays in schema.\n\n**Step 3 — validate.**\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Farch-index\u002Fscripts\u002Fvalidate_index.py .arch\u002Findex.json --repo-root .\n```\n\nExit zero proceeds. Non-zero loops back to step 2 with the exact violation lines,\nbounded to three rounds, then stops and reports the remaining violations. **If the\nvalidator cannot be executed at all, that is a BLOCK, not a pass** — there is no prose\nfallback, because a fallback an agent can narrate is the failure this gate exists to\nremove.\n\n**Step 4 — write the derived artifacts,** using `templates\u002F`. Write\n`.arch\u002FARCH-EVIDENCE.md` capturing the command, the exit code, and the per-check table.\nWrite `.arch\u002FINDEX.md` for humans. Emit the Mermaid `C4Container` source and render it\nthrough `design:mermaid` (beautiful-mermaid), never the standard Mermaid CDN.\n\n**Step 5 — advance the pin.** Write `git rev-parse HEAD` to `.arch\u002Fpinned-commit`\n**only after the validator has exited zero.** A failed run leaves the previous pin\nuntouched, so a broken index can never present itself as current.\n\n### `drift`\n\n**Deterministic first, and often only.**\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Farch-index\u002Fscripts\u002Farch_drift.py --repo-root .\n```\n\nThe script reads the pin, diffs it against HEAD with `--name-status` and `--numstat`,\ntests every primary path, and writes `.arch\u002FDRIFT.json` with a per-component verdict of\n`REMOVE`, `PRUNE`, or `KEEP`, plus the files added outside every claimed path. It\nresolves the gate itself:\n\n| Gate | Meaning | What to do |\n|---|---|---|\n| `NONE` | Every verdict is `KEEP` and nothing unclaimed was added | Report `NO DRIFT` and stop. **Zero model calls.** |\n| `FULL_REBUILD` | More than 300 files or 50,000 lines changed since the pin | A rebuild is cheaper than a patch. Run `build`. |\n| `INCREMENTAL` | Anything else | Continue below. |\n\nIf the script exits 2, the scan could not run. That is never `NO DRIFT`; report why.\n\nIn the `INCREMENTAL` case the model receives `DRIFT.json` as **settled fact** and\nanswers exactly one open question: do the unclaimed added files constitute a component\nthat should exist? Apply the verdicts without re-deciding them, preserve surviving\ncomponent identifiers, prune connections to removed components, then validate and\nadvance the pin as in `build` steps 3 through 5.\n\nNever preserve a component whose code is gone. See the deletion-first rule in\n`references\u002Fcomponent-rules.md`.\n\n### `check`\n\nRuns the validator alone and writes `.arch\u002FARCH-EVIDENCE.md`. Makes no model calls and\nnever writes `index.json`. This is what `\u002Fteam verify` and `\u002Fteam ship` call.\n\n## Artifacts\n\n| Path | Committed | Notes |\n|---|---|---|\n| `.arch\u002Findex.json` | Yes | The contract. Diffable history is the point. |\n| `.arch\u002Fpinned-commit` | Yes | Advanced only on validator exit 0. |\n| `.arch\u002FINDEX.md` | Yes | Derived; regenerated every build. |\n| `.arch\u002FARCH-EVIDENCE.md` | Yes | The captured gate. |\n| `.arch\u002FANALYSIS.md` | Yes | Fallback path only; cached against the pin. |\n| `.arch\u002FDRIFT.json` | No | Derived scratch; gitignored. |\n\n`.arch\u002F` is committed deliberately. An untracked file cannot be diffed across runs, does\nnot exist in the clean checkout that `\u002Fteam ship` verifies from, and would be silently\nerased by `git clean`, reverting every consumer to a degraded path without saying so.\n\n## Scope limits, stated plainly\n\n**Structural drift only.** A commit that swaps a component's datastore, rewrites it end\nto end, or violates a declared boundary entirely inside already-claimed paths returns\n`NO DRIFT`. Every artifact carries these words, and no consumer may drop them.\n\n**Currency is caller-enforced.** No script compares the pin to HEAD; the caller does,\nper `team:architecture.md` step 2. An index behind HEAD is `STALE`, and claims resting\non it are `UNVERIFIED` rather than clean.\n\n**The read budget is not enforced.** There is no transcript API and no way to count a\nsubagent's tool calls from outside it. `references\u002Ftool-frugality.md` is guidance, and\nno artifact reports it as verified.\n\n**Monorepos do not fit.** The three-to-eight ladder forces distinct concerns to merge in\na repository with dozens of packages. One index per workspace package plus a root index\nwhose components are the packages is the eventual answer, and is out of scope for 1.0.\n\n## Files\n\n| File | Purpose |\n|---|---|\n| `references\u002Fschema.md` | The contract and the twelve checks |\n| `references\u002Farch-plan.md` | The forward plan contract, and why declare-then-verify works |\n| `references\u002Fcomponent-rules.md` | What qualifies, naming, counting, merging |\n| `references\u002Ftool-frugality.md` | The read budget for the fallback crawl |\n| `references\u002Fanalysis-taxonomy.md` | The seven-section exploration skeleton |\n| `scripts\u002Fvalidate_index.py` | The gate. Twelve checks, stdlib only, exit 0 or 1 |\n| `scripts\u002Farch_drift.py` | Deterministic drift scan. Zero model calls |\n| `scripts\u002Fverify_arch_plan.py` | Forward plan: `--declare` before a build, `--verify` after |\n| `scripts\u002Frepo_tree.py` | Filtered, depth-adaptive tree |\n| `scripts\u002Frun_fixtures.sh` | Regression suite for the validator |\n| `templates\u002F` | `INDEX.md` and `ARCH-EVIDENCE.md` skeletons |\n\nRegression gate for any change to the validator:\n\n```bash\nbash skills\u002Farch-index\u002Fscripts\u002Frun_fixtures.sh\n```\n\n## Attribution\n\nThe design this skill implements is adapted from **devildev** by lak7, licensed\n**Apache-2.0**, \u003Chttps:\u002F\u002Fgithub.com\u002Flak7\u002Fdevildev>. No upstream source code was copied;\nwhat was taken is prompt text, one schema idea, and four procedures.\n\n| Borrowed | Upstream source |\n|---|---|\n| Component-to-code ownership taxonomy | `prompts\u002FReverseArchitecture.ts` |\n| Component rules, count ladder, naming and purpose formulas, anti-patterns | `prompts\u002FReverseArchitecture.ts:278-371`, `:493-507` |\n| Deletion-first reconciliation | `prompts\u002FReverseArchitecture.ts:578-641` |\n| Two-stage explore-then-synthesise split | `actions\u002Freverse-architecture.ts:696-805` |\n| Tool-frugality read budget | `actions\u002Freverse-architecture.ts:715-749` |\n| Seven-section analysis taxonomy and character cap | `actions\u002Freverse-architecture.ts:771-805` |\n\nPer-file modification notices required by Apache-2.0 § 4(b) are recorded at the foot of\neach reference file. The most significant modification is that upstream stated its\nexact-path-matching rule emphatically in prose, with a validation checklist of unchecked\nboxes, and never verified a single path in roughly 1,300 lines of code; here that rule\nis `scripts\u002Fvalidate_index.py` checks 5 and 6, with a process exit code. Also not\ncarried across: all presentation fields, numeric confidence bands, the celebrity-persona\nframing, the framework-only admission gate, and the credits economy.\n","# arch-index\n\nProduces one artifact: `.arch\u002Findex.json`, a commit-pinned map from each architectural\ncomponent of a repository to the real directories and files that implement it. Every\nclaimed path is verified to exist by a script with an exit code, not by a promise in a\nprompt.\n\n**Guiding principle — Paths or it does not exist.** The protocol that consumers follow\nis `team:architecture.md`. This file is the producer.\n\n## What this skill is for, and what it is not for\n\nIt is for answering, mechanically: which files implement this component, and has that\nstopped being true.\n\nIt is not a design tool. The index is **reverse-only** — it describes code that exists\non disk at a known commit. There is no mode for specifying a system that has not been\nbuilt, because the central check is that every path resolves. Forward design work\nbelongs in `\u002Fteam think`, the `SI-*-Design` councils, or\n`\u002Futil:create-architecture-documentation`.\n\nIt is not a codebase-mapping tool either, and it should not act like one. When\n`.planning\u002Fcodebase\u002F` exists, `gsd-map-codebase` has already produced seven structured\ndocuments covering the stack, integrations, architecture, structure, conventions,\ntesting, and concerns. This skill **reads those and skips crawling entirely.** Crawling\na repository that has already been mapped is the exact waste this skill was designed to\navoid, and duplicating `gsd-map-codebase` would make it a worse version of a tool that\nalready exists.\n\nIt is not a work-allocation mechanism. The index sits at C4-Container altitude, so a\nsingle component can own thousands of files. Nothing derives per-agent file ownership\nfrom it.\n\n## Modes\n\nResolve the mode from the invocation. When none is given, use `drift` if\n`.arch\u002Fpinned-commit` exists and `build` if it does not.\n\n### `build`\n\n**Step 1 — find prior art before doing any work.** In this order:\n\n1. If `.planning\u002Fcodebase\u002F` exists, read `ARCHITECTURE.md`, `STRUCTURE.md`,\n   `STACK.md`, and `INTEGRATIONS.md` from it. Set `source` to `gsd-codebase-map`. Skip\n   to step 2. Do not crawl.\n2. Otherwise, if gitnexus is available for this repository, call\n   `mcp__gitnexus__context` and `mcp__gitnexus__group_list` for a deterministic view of\n   module grouping. Set `source` to `gitnexus`. Skip to step 2.\n\n   A gitnexus index that is behind HEAD is still usable here, and `list_repos` reports\n   exactly how far behind it is. Record that number in `ecosystem.sourceCommitsBehind`.\n   The reason staleness is tolerable is structural rather than optimistic: gitnexus\n   informs *component grouping*, while every path the index claims is checked by the\n   validator against the working tree. A path that moved since gitnexus was indexed\n   fails check 5 and blocks the build. Stale grouping advice can therefore make the\n   decomposition slightly dated, but it cannot introduce a path that does not exist.\n   Still run `scripts\u002Frepo_tree.py` for the authoritative current tree.\n3. Otherwise crawl. Run `scripts\u002Frepo_tree.py --repo-root .` for a filtered,\n   depth-adaptive tree, then perform one exploration pass under\n   `references\u002Ftool-frugality.md`, writing `.arch\u002FANALYSIS.md` against the seven-section\n   skeleton in `references\u002Fanalysis-taxonomy.md` under its ten-thousand-character cap.\n   Set `source` to `crawl`.\n\nIf the tree came back truncated, say so at the top of the analysis and set\n`ecosystem.treeTruncated` to `true` in the index. A truncated tree disables `drift`\nmode entirely, by design.\n\n**Step 2 — synthesise, with no tool use.** Read only the gathered summary plus the\ntree. Emit `.arch\u002Findex.json` per `references\u002Fschema.md`, bound by\n`references\u002Fcomponent-rules.md`. This stage does not explore; separating exploration\nfrom formatting is why the structured output stays in schema.\n\n**Step 3 — validate.**\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Farch-index\u002Fscripts\u002Fvalidate_index.py .arch\u002Findex.json --repo-root .\n```\n\nExit zero proceeds. Non-zero loops back to step 2 with the exact violation lines,\nbounded to three rounds, then stops and reports the remaining violations. **If the\nvalidator cannot be executed at all, that is a BLOCK, not a pass** — there is no prose\nfallback, because a fallback an agent can narrate is the failure this gate exists to\nremove.\n\n**Step 4 — write the derived artifacts,** using `templates\u002F`. Write\n`.arch\u002FARCH-EVIDENCE.md` capturing the command, the exit code, and the per-check table.\nWrite `.arch\u002FINDEX.md` for humans. Emit the Mermaid `C4Container` source and render it\nthrough `design:mermaid` (beautiful-mermaid), never the standard Mermaid CDN.\n\n**Step 5 — advance the pin.** Write `git rev-parse HEAD` to `.arch\u002Fpinned-commit`\n**only after the validator has exited zero.** A failed run leaves the previous pin\nuntouched, so a broken index can never present itself as current.\n\n### `drift`\n\n**Deterministic first, and often only.**\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Farch-index\u002Fscripts\u002Farch_drift.py --repo-root .\n```\n\nThe script reads the pin, diffs it against HEAD with `--name-status` and `--numstat`,\ntests every primary path, and writes `.arch\u002FDRIFT.json` with a per-component verdict of\n`REMOVE`, `PRUNE`, or `KEEP`, plus the files added outside every claimed path. It\nresolves the gate itself:\n\n| Gate | Meaning | What to do |\n|---|---|---|\n| `NONE` | Every verdict is `KEEP` and nothing unclaimed was added | Report `NO DRIFT` and stop. **Zero model calls.** |\n| `FULL_REBUILD` | More than 300 files or 50,000 lines changed since the pin | A rebuild is cheaper than a patch. Run `build`. |\n| `INCREMENTAL` | Anything else | Continue below. |\n\nIf the script exits 2, the scan could not run. That is never `NO DRIFT`; report why.\n\nIn the `INCREMENTAL` case the model receives `DRIFT.json` as **settled fact** and\nanswers exactly one open question: do the unclaimed added files constitute a component\nthat should exist? Apply the verdicts without re-deciding them, preserve surviving\ncomponent identifiers, prune connections to removed components, then validate and\nadvance the pin as in `build` steps 3 through 5.\n\nNever preserve a component whose code is gone. See the deletion-first rule in\n`references\u002Fcomponent-rules.md`.\n\n### `check`\n\nRuns the validator alone and writes `.arch\u002FARCH-EVIDENCE.md`. Makes no model calls and\nnever writes `index.json`. This is what `\u002Fteam verify` and `\u002Fteam ship` call.\n\n## Artifacts\n\n| Path | Committed | Notes |\n|---|---|---|\n| `.arch\u002Findex.json` | Yes | The contract. Diffable history is the point. |\n| `.arch\u002Fpinned-commit` | Yes | Advanced only on validator exit 0. |\n| `.arch\u002FINDEX.md` | Yes | Derived; regenerated every build. |\n| `.arch\u002FARCH-EVIDENCE.md` | Yes | The captured gate. |\n| `.arch\u002FANALYSIS.md` | Yes | Fallback path only; cached against the pin. |\n| `.arch\u002FDRIFT.json` | No | Derived scratch; gitignored. |\n\n`.arch\u002F` is committed deliberately. An untracked file cannot be diffed across runs, does\nnot exist in the clean checkout that `\u002Fteam ship` verifies from, and would be silently\nerased by `git clean`, reverting every consumer to a degraded path without saying so.\n\n## Scope limits, stated plainly\n\n**Structural drift only.** A commit that swaps a component's datastore, rewrites it end\nto end, or violates a declared boundary entirely inside already-claimed paths returns\n`NO DRIFT`. Every artifact carries these words, and no consumer may drop them.\n\n**Currency is caller-enforced.** No script compares the pin to HEAD; the caller does,\nper `team:architecture.md` step 2. An index behind HEAD is `STALE`, and claims resting\non it are `UNVERIFIED` rather than clean.\n\n**The read budget is not enforced.** There is no transcript API and no way to count a\nsubagent's tool calls from outside it. `references\u002Ftool-frugality.md` is guidance, and\nno artifact reports it as verified.\n\n**Monorepos do not fit.** The three-to-eight ladder forces distinct concerns to merge in\na repository with dozens of packages. One index per workspace package plus a root index\nwhose components are the packages is the eventual answer, and is out of scope for 1.0.\n\n## Files\n\n| File | Purpose |\n|---|---|\n| `references\u002Fschema.md` | The contract and the twelve checks |\n| `references\u002Farch-plan.md` | The forward plan contract, and why declare-then-verify works |\n| `references\u002Fcomponent-rules.md` | What qualifies, naming, counting, merging |\n| `references\u002Ftool-frugality.md` | The read budget for the fallback crawl |\n| `references\u002Fanalysis-taxonomy.md` | The seven-section exploration skeleton |\n| `scripts\u002Fvalidate_index.py` | The gate. Twelve checks, stdlib only, exit 0 or 1 |\n| `scripts\u002Farch_drift.py` | Deterministic drift scan. Zero model calls |\n| `scripts\u002Fverify_arch_plan.py` | Forward plan: `--declare` before a build, `--verify` after |\n| `scripts\u002Frepo_tree.py` | Filtered, depth-adaptive tree |\n| `scripts\u002Frun_fixtures.sh` | Regression suite for the validator |\n| `templates\u002F` | `INDEX.md` and `ARCH-EVIDENCE.md` skeletons |\n\nRegression gate for any change to the validator:\n\n```bash\nbash skills\u002Farch-index\u002Fscripts\u002Frun_fixtures.sh\n```\n\n## Attribution\n\nThe design this skill implements is adapted from **devildev** by lak7, licensed\n**Apache-2.0**, \u003Chttps:\u002F\u002Fgithub.com\u002Flak7\u002Fdevildev>. No upstream source code was copied;\nwhat was taken is prompt text, one schema idea, and four procedures.\n\n| Borrowed | Upstream source |\n|---|---|\n| Component-to-code ownership taxonomy | `prompts\u002FReverseArchitecture.ts` |\n| Component rules, count ladder, naming and purpose formulas, anti-patterns | `prompts\u002FReverseArchitecture.ts:278-371`, `:493-507` |\n| Deletion-first reconciliation | `prompts\u002FReverseArchitecture.ts:578-641` |\n| Two-stage explore-then-synthesise split | `actions\u002Freverse-architecture.ts:696-805` |\n| Tool-frugality read budget | `actions\u002Freverse-architecture.ts:715-749` |\n| Seven-section analysis taxonomy and character cap | `actions\u002Freverse-architecture.ts:771-805` |\n\nPer-file modification notices required by Apache-2.0 § 4(b) are recorded at the foot of\neach reference file. The most significant modification is that upstream stated its\nexact-path-matching rule emphatically in prose, with a validation checklist of unchecked\nboxes, and never verified a single path in roughly 1,300 lines of code; here that rule\nis `scripts\u002Fvalidate_index.py` checks 5 and 6, with a process exit code. Also not\ncarried across: all presentation fields, numeric confidence bands, the celebrity-persona\nframing, the framework-only admission gate, and the credits economy.\n",{"name":5,"domain":54,"version":55,"description":7},"engineering","1.0.0","main","cd89f013b617f7eea982a9bbc60e992974fda2aa","skills\u002Farch-index",15,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco","https:\u002F\u002Fcocoresearch.org\u002Fcoco\u002F","https:\u002F\u002Fgithub.com\u002Fcoco-research","https:\u002F\u002Fcocoresearch.org","Local-first, open-source tools for people who work with AI agents: governed memory, agents that check their own work, and data that stays on your machine.","2026-11-10 10:38:36",{"source":8,"skillName":5,"commands":67,"prompt":77,"zipAvailable":78},[68,71,74],{"tool":69,"command":70},"npx","npx skills add coco-research\u002Fcoco --skill arch-index",{"tool":72,"command":73},"bunx","bunx skills add coco-research\u002Fcoco --skill arch-index",{"tool":75,"command":76},"pnpm","pnpm dlx skills add coco-research\u002Fcoco --skill arch-index","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002FSKILL.md",true,[80,86,90,94,98,102,108,112,116,120,124,128,132,136,140,144,148,154],{"path":81,"resourceType":82,"resourceTypeDesc":83,"sizeBytes":84,"sourceHtmlUrl":85},"references\u002Fanalysis-taxonomy.md","REFERENCE","参考文档",4817,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Freferences\u002Fanalysis-taxonomy.md",{"path":87,"resourceType":82,"resourceTypeDesc":83,"sizeBytes":88,"sourceHtmlUrl":89},"references\u002Farch-plan.md",8054,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Freferences\u002Farch-plan.md",{"path":91,"resourceType":82,"resourceTypeDesc":83,"sizeBytes":92,"sourceHtmlUrl":93},"references\u002Fcomponent-rules.md",6983,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Freferences\u002Fcomponent-rules.md",{"path":95,"resourceType":82,"resourceTypeDesc":83,"sizeBytes":96,"sourceHtmlUrl":97},"references\u002Fschema.md",9366,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Freferences\u002Fschema.md",{"path":99,"resourceType":82,"resourceTypeDesc":83,"sizeBytes":100,"sourceHtmlUrl":101},"references\u002Ftool-frugality.md",4759,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Freferences\u002Ftool-frugality.md",{"path":103,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":106,"sourceHtmlUrl":107},"scripts\u002Farch_drift.py","SCRIPT","脚本",8638,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Farch_drift.py",{"path":109,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":110,"sourceHtmlUrl":111},"scripts\u002Ffixtures\u002Fbad-invariants.json",1638,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fbad-invariants.json",{"path":113,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":114,"sourceHtmlUrl":115},"scripts\u002Ffixtures\u002Fbad-paths.json",1933,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fbad-paths.json",{"path":117,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":118,"sourceHtmlUrl":119},"scripts\u002Ffixtures\u002Fgood.json",3322,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fgood.json",{"path":121,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":122,"sourceHtmlUrl":123},"scripts\u002Ffixtures\u002Fplan-bad-shape.json",1292,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fplan-bad-shape.json",{"path":125,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":126,"sourceHtmlUrl":127},"scripts\u002Ffixtures\u002Fplan-good.json",1204,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fplan-good.json",{"path":129,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":130,"sourceHtmlUrl":131},"scripts\u002Ffixtures\u002Fplan-unbuilt.json",1343,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Ffixtures\u002Fplan-unbuilt.json",{"path":133,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":134,"sourceHtmlUrl":135},"scripts\u002Frepo_tree.py",7541,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Frepo_tree.py",{"path":137,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":138,"sourceHtmlUrl":139},"scripts\u002Frun_fixtures.sh",7353,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Frun_fixtures.sh",{"path":141,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":142,"sourceHtmlUrl":143},"scripts\u002Fvalidate_index.py",14684,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Fvalidate_index.py",{"path":145,"resourceType":104,"resourceTypeDesc":105,"sizeBytes":146,"sourceHtmlUrl":147},"scripts\u002Fverify_arch_plan.py",10407,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Fscripts\u002Fverify_arch_plan.py",{"path":149,"resourceType":150,"resourceTypeDesc":151,"sizeBytes":152,"sourceHtmlUrl":153},"templates\u002FARCH-EVIDENCE.md.template","OTHER","其他",2060,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Ftemplates\u002FARCH-EVIDENCE.md.template",{"path":155,"resourceType":150,"resourceTypeDesc":151,"sizeBytes":156,"sourceHtmlUrl":157},"templates\u002FINDEX.md.template",1307,"https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fskills\u002Farch-index\u002Ftemplates\u002FINDEX.md.template",[159,184,194,210,245,264],{"skillNo":160,"name":161,"nameZh":6,"description":162,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":163,"categories":164,"keywords":166,"publishedAt":49},"01M71KEJXF9KMT0BWCMRSRFA4M","trading-super-intelligence","Trading Super Intelligence Team. Named real-world personas across 8 cells. Built local-first (LM Studio) and validator-gated. Illustrative composites; see DISCLAIMER.md.","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Ftrading\u002FSKILL.md",[165],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183],"super","trading super intelligence","intelligence","trading","validator","personas","local","composites","studio","team","across","world","first","gated","2026","lm studio","local-first",{"skillNo":185,"name":186,"nameZh":6,"description":187,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":188,"categories":189,"keywords":191,"publishedAt":49},"01M71KEFWS3CMAP7DF1P0FWTDE","strategy-super-intelligence","Strategy Super Intelligence Team. Named real-world personas across 7 cells. Built local-first (LM Studio) and validator-gated. Illustrative composites; see DISCLAIMER.md.","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Fstrategy\u002FSKILL.md",[190],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[192,167,193,169,171,172,173,174,175,176,177,178,179,180,181,182,183],"strategy","strategy super intelligence",{"skillNo":195,"name":196,"nameZh":6,"description":197,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":198,"categories":199,"keywords":204,"publishedAt":49},"01M71KED56ZEPK2QE84QSQPMT0","risk-compliance-super-intelligence","Risk & Compliance (GRC) Super Intelligence Team. Named real-world personas across 7 cells. Built local-first (LM Studio) and validator-gated. Illustrative composites; see DISCLAIMER.md.","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Frisk-compliance\u002FSKILL.md",[200,201],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":202,"categoryName":203,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000037","律师与法务",[205,167,206,207,208,169,171,172,173,174,209,175,176,177,178,179,180,181,182,183],"risk","risk compliance super intelligence","compliance","risk compliance","grc",{"skillNo":211,"name":212,"nameZh":6,"description":213,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":214,"categories":215,"keywords":229,"publishedAt":49},"01M71KE9A2FWSC74VA96W1EZSN","product-design-super-intelligence","Your product + design brain trust and decision partner. Named personas across 8 cells spanning product strategy, discovery\u002Fresearch, growth\u002Fmetrics, design foundations\u002Fusability, design leadership\u002Fcraft, UX\u002Fcontent, design systems, and a…","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Fproduct-design\u002FSKILL.md",[216,219,222,223,226],{"categoryNo":217,"categoryName":218,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":220,"categoryName":221,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000021","资料检索与研究",{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":224,"categoryName":225,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000031","产品经理",{"categoryNo":227,"categoryName":228,"taxonomyType":26,"taxonomyTypeDesc":27},"01JBSEEDCAT000000000000032","设计师",[230,231,169,232,167,233,234,235,236,237,238,176,177,239,240,181,241,242,243,244],"design","product","product design","product design super intelligence","foundations","product-design","decision","design systems","usability","leadership","ux","brain","craft","metrics","trust",{"skillNo":246,"name":247,"nameZh":6,"description":248,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":249,"categories":250,"keywords":253,"publishedAt":49},"01M71KE6H9FPSD9TECYFBPXZ8Y","legal-ip-super-intelligence","Your legal, IP, and regulatory brain trust. 31 named personas across 7 cells covering patent strategy, tech licensing, regulatory litigation, AI governance, open-source compliance, privacy law, and antitrust.","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Flegal-ip\u002FSKILL.md",[251,252],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":202,"categoryName":203,"taxonomyType":26,"taxonomyTypeDesc":27},[254,167,255,169,256,257,244,258,172,259,260,261,192,176,177,262,207,263,181,241],"ip","legal ip super intelligence","legal","sep","tech","litigation","ai governance","open","open-source","governance",{"skillNo":265,"name":266,"nameZh":6,"description":267,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":268,"categories":269,"keywords":271,"publishedAt":49},"01M71KE36QRY610V6YVB9JWRND","healthcare-life-sciences-super-intelligence","Your healthcare and life sciences brain trust. 25 named personas across 7 cells covering drug discovery, medtech, biotech strategy, health policy, genomics, clinical trials, and healthcare economics.","https:\u002F\u002Fgithub.com\u002Fcoco-research\u002Fcoco\u002Fblob\u002Fmain\u002Fsystems\u002Fsuperintelligence\u002Fhealthcare-life-sciences\u002FSKILL.md",[270],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[272,167,273,274,169,257,244,275,276,172,277,278,192,176,177,279,280,281,282,181],"healthcare","healthcare life sciences super intelligence","life","economics","discovery","policy","clinical","clinical trials","health","trials","genomics"]