docs-sync
Skill77 starsChanged 6 days agoMITView on GitHub
Keep docs and config examples coherent after a change - ARCHITECTURE.md, README, GOAL.md, MODELS.md, ROADMAP.md (ledger rows), imp.conf.example, CHANGELOG, MISSION_JOURNAL, "README says X but code does Y". Not layers/lint (docs-layers), code audits (codebase-audit), perf baseline (benchmark-cuda), agent memory.
46 other projects use a version of this.
Adopted by
Most recently taken up by
- julianosirtori/skillstheir own version, first seen Nov 8, 2026
- voyager67/AgentFieldkittheir own version, first seen Nov 8, 2026
- uwuclxdy/agenticattheir own version, first seen Nov 8, 2026
- sunholo-data/ailangtheir own version, first seen Nov 6, 2026
- jackneil/claude-jackedtheir own version, first seen Nov 5, 2026
- yasik/agentlanetheir own version, first seen Nov 1, 2026
Used with
Projects with it also use tdd, code-review and 2 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
GoogleSkilltdd
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
Also in 4 projects that use this
- GoogleSkill
code-review
Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.
Also in 5 projects that use this
MicrosoftSkillhandoff
Compact the current conversation into a handoff document for another agent to pick up.
Also in 4 projects that use this
- MicrosoftSkill
release
Use ONLY when the user explicitly asks to cut/publish a Huabu desktop release or tag a version. Performs an outward action — bumps the version in apps/desktop/package.json, commits + annotated-tags vX.Y.Z, and pushes the branch + tag. That tag push auto-builds the macOS/Windows installers and publishes a GitHub Release. Provides `status`, `bump`, `tag-push`, and a one-command `run` (bump → tag-push).
Also in 3 projects that use this
Other versions46 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 46
- openai/openai-agents-python★ 31ksame name
- redis/node-redis★ 18ksame name
- redis/ioredis★ 15ksame name
- fossasia/visdom★ 10ksame name
- Kiln-AI/Kiln★ 5.3ksame name
- danielmeppiel/apm★ 4.2ksame name
- microsoft/apm★ 4.2ksame name
- openai/openai-agents-js★ 4ksame name
- pnp/powershell★ 912same name
- strands-agents/box★ 503same name
36 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- tests/CLAUDE.mdCLAUDE.md
- add-model-archSkill
- benchmark-cudaSkill
- building-and-testingSkill
- check-degenerationSkill
- codebase-auditSkill
- code-graphSkill
- docs-layersSkill
- find-stubsSkill
- numerics-bisectSkill
- quant-formatsSkill
- server-apiSkill
- shipping-prsSkill
- sm120-cuda-expertSkill
The file
Files
Docs sync - imp
Hard rules
| # | Rule |
| 1 | imp.conf.example MUST match the parser: every key is a B/I/F/S("...") binder in src/runtime/config.cpp (surface in src/core/config/*.h); a dropped key logs imp.conf: unknown key (stale q4k_imma_enabled after #624). Add/remove/rename = config.h, config.cpp, imp.conf.example together, real default. A bogus --set key fails at start (no such key). |
| 2 | Numbers are commit-anchored; docs/PERF.md owns them. tests/perf_baseline.json is the gate (8%/8%/10% own_peak_mb, scripts/verify.sh); the docs/PERF.md perf block is GENERATED (scripts/sync_docs.py, <!-- PERF:BEGIN -->; --check blocks). docs/BENCHMARKS.md rows: date, commit SHA, CUDA, model, quant, metric, value, exact command. |
| 3 | Verify before you claim: grep the tree; half the staleness reports are already fixed (the "GOAL.md lists H100" flag was one). |
| 4 | Env vars are not config: the engine reads only IMP_CONFIG, IMP_DETERMINISTIC, IMP_FMHA_FA2, IMP_WORKER_TIMING, IMP_SPEC_TRACE, IMP_JUMP_TRACE, IMP_PPL_DUMP; config is imp.conf / --config / --set. The container surface is IMP_CONFIG + IMP_SET (#1823, docs/DEPLOYMENT.md "From a container"); the 19 legacy IMP_* entrypoint names are frozen and get no new siblings. |
| 5 | Three blocking doc gates, in hooks and Build: scripts/docs_lint.py, scripts/sync_docs.py --check, scripts/check_doc_citations.py . (all living docs since #1783; a TU split or a de-prose compaction moves line numbers). |
| 6 | No prose: a paragraph without a number, path or decision is deleted (#1802/#1804 sweeps). |
The doc set
| Doc | Owns | Touch when |
docs/internals/ARCHITECTURE.md | canonical narrative | a refactor changes structure or data flow |
docs/internals/MEMORY.md, root AUDIT.md, docs/audit/AUDIT_ARCH_2026_07_29.md | memory subsystem (tiers, allocators, invariants I1-I7), running findings log (incl. REFUTED), 07-29 architecture audit | ownership, lifetime, capacity, VRAM behaviour |
README.md | pitch + quickstart; no CI gate numbers | supported models or build steps change |
docs/PERF.md | every number (partly generated) | any perf number changes |
docs/GOAL.md | north star (Qwen3-14B Q6_K @ctx2048 -> 175 tok/s) + hardware scope | goal or scope changes |
docs/BENCHMARKS.md | reproducible SHA-anchored table; **Toolchain (current: \vX.Y.Z`):**line parsed bycheck-release.sh` | a publishable number; a release |
docs/FEATURES.md, docs/LIMITATIONS.md, docs/DESIGN_DECISIONS.md | what exists / what does not / why | feature ships, limitation opens or closes, decision made or reversed |
docs/ROADMAP.md | verdict ledgers (fact + number + decision per row), citations drift-gated by check_doc_citations.py (#1772) | a lever is measured, shipped, refuted or closed; investigation goes to docs/plans/YYYY-MM-DD-<topic>.md |
docs/plans/ | records with closure marks (docs-layers) | an item closes |
tests/perf_baseline*.json | perf + peak-VRAM gate | intentional perf/VRAM move (scripts/gen_perf_baseline.sh) |
docs/MODELS.md, docs/DEPLOYMENT.md, docs/API.md, docs/QUICKSTART.md | supported models; container/compose; HTTP contract (request tracing, priority); worked example (kekzle/Qwen3.8-27B-NVFP4-vllm) | new arch/model/quant; new flag or env; new endpoint field |
docs/MISSION_JOURNAL.md, docs/scoreboard.tsv | mission log, competitive scoreboard | competitive result changes |
CHANGELOG.md | 1-3 lines per entry, fact + number + (#PR), merged into the existing ### Added/Changed/Fixed block (shipping-prs) | user-visible change |
docs/audit/ | audit reports and ledgers (codebase-audit) | an audit pass |
.claude/skills/*/SKILL.md, CLAUDE.md tree | L3 agent routing; per-dir CLAUDE.md <= 800 tokens, no perf numbers | a workflow, gate or trap changes |
When a change lands
| Change | Sync |
| Perf moved intentionally | gen_perf_baseline.sh, BENCHMARKS.md row, say so in the PR; measure per benchmark-cuda first |
| Config key added/removed/renamed | config.h + config.cpp + imp.conf.example; a new auto-default also needs an imp-cli --bench pin (apply_config_pins in tools/imp-cli/args.cpp) or the baseline bakes it in |
| New arch / model / quant | docs/MODELS.md (+ README if headline) |
| New endpoint field, header or env | docs/API.md, docs/DEPLOYMENT.md, tests/api/mock_server.py mirrors the contract |
| Structural refactor | ARCHITECTURE.md if the narrative broke, then python3 scripts/check_doc_citations.py . (the #1782 scheduler split cost a CI roundtrip on a ROADMAP.md citation) |
| A lever measured (shipped or refuted) | docs/ROADMAP.md ledger row; harness in tools/analysis/; record in docs/plans/ if the investigation is longer than three lines |
| Any L0-L2 doc edited | bump verified:/commit: in its header (docs-layers) |
Common mistakes
- A
BENCHMARKS.mdnumber from a cold single shot (benchmark-cuda STOP list). - A skill or
CLAUDE.mdcarrying a count or runtime that nothing checks (docs-layers rule 3b).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.