Not in the top 1,000 by GitHub stars

WoJiSama/skill-based-architecture

A meta-skill that produces skills. Point it at any codebase and it distills the project's rules, workflows, and hard-won lessons into a dedicated skills/<name>/ directory — a project skill that becomes the single source of truth every AI agent (Cursor, Claude Code, Codex, Windsurf, Gemini) consults before every task.

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

e5d2da58ec68 → e5d2da58ec68

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e5d2da58ec68

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit e5d2da58ec6812b4615d39869f630cfd20fd58a5

git clone --filter=blob:none --no-checkout https://github.com/WoJiSama/skill-based-architecture.git .hotskills/wojisama-skill-based-architecture && git -C .hotskills/wojisama-skill-based-architecture checkout --detach e5d2da58ec6812b4615d39869f630cfd20fd58a5 # Copy each selected skill directory from .hotskills/wojisama-skill-based-architecture into .claude/skills/<skill-name>/
git clone --filter=blob:none --no-checkout https://github.com/WoJiSama/skill-based-architecture.git .hotskills/wojisama-skill-based-architecture && git -C .hotskills/wojisama-skill-based-architecture checkout --detach e5d2da58ec6812b4615d39869f630cfd20fd58a5 # Copy each selected skill directory from .hotskills/wojisama-skill-based-architecture into .agents/skills/<skill-name>/
git clone --filter=blob:none --no-checkout https://github.com/WoJiSama/skill-based-architecture.git .hotskills/wojisama-skill-based-architecture && git -C .hotskills/wojisama-skill-based-architecture checkout --detach e5d2da58ec6812b4615d39869f630cfd20fd58a5 # Copy each selected skill directory from .hotskills/wojisama-skill-based-architecture into .cursor/skills/<skill-name>/
git clone --filter=blob:none --no-checkout https://github.com/WoJiSama/skill-based-architecture.git .hotskills/wojisama-skill-based-architecture && git -C .hotskills/wojisama-skill-based-architecture checkout --detach e5d2da58ec6812b4615d39869f630cfd20fd58a5 # Copy each selected skill directory from .hotskills/wojisama-skill-based-architecture into .gemini/skills/<skill-name>/
619GitHub stars
-1star in 7 days
1skill in this repository
0commits in 30 days

Heat

0.2
  • Star growth, last 7 daysNot available
  • Star growth, last 24 hoursNot available
  • Rank movement, last 7 daysNot available
  • Commits, last 30 daysNot available
  • Accounts asking about it through MCP, last 7 daysNot available
  • Hot votes this weekNot available

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Oct 25, 2026Oct 31, 2026622

30-day change not yet available: not enough history.

Individual skills in this repository

How it works

skill-based-architecture

When your agent uses it

This skill should be used when the user asks to "organize the project rules", "clean up scattered documentation", "把规则迁移到 skills 目录", "优化 skill 路由", "提高 description 命中率", or "减少薄壳重复维护". Activate when a SKILL.md is too large, rules are duplicated across agent entry files, task routing or trigger_examples miss natural user language, or templates / thin shells / validation scripts need drift-resistant maintenance.

What it loads into context

Frontmatter

116 tokens

Always available to the agent

SKILL.md

2,846 tokens

When the skill is activated

Reference documents

308,910 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. When to Use
  2. Progressive Rigor
  3. Evidence-Selected Structure
  4. Core Principles
  5. Common Pitfalls
  6. Content Classification
  7. Multi-Skill & Composition
  8. Resources

Scripts executed separately

  • .claude/hooks/session-start
  • examples/simple-repo/repo/package.json
  • examples/simple-repo/repo/src/cart.js
  • scripts/check-all.sh
  • scripts/check-live-agent-journey.sh
  • scripts/check-migration-evidence.sh
  • scripts/check-self-scenarios.sh
  • scripts/check-self-shells.sh
  • scripts/check-template-hooks.sh
  • scripts/check-upstream-changes.sh
  • scripts/check-upstream-supersedes.sh
  • scripts/check-validation-contract.sh
  • scripts/scaffold-downstream.sh
  • scripts/skill-asset
  • scripts/sync-self-shells.sh
  • scripts/test.js
  • templates/hooks/session-start
  • templates/hooks/workflow-state
  • templates/skill/scripts/audit-orphans.sh
  • templates/skill/scripts/check-version-conformance.sh
  • templates/skill/scripts/route-health.sh
  • templates/skill/scripts/route-reachability.sh
  • templates/skill/scripts/smoke-test.sh
  • templates/skill/scripts/sync-routing.sh
  • templates/skill/scripts/sync-vendor.sh
  • templates/skill/scripts/upstream-status.sh

Files

  • .claude-plugin/marketplace.jsonasset820 bytes
  • .claude-plugin/plugin.jsonasset524 bytes
  • .claude/hooks/session-startruns935 bytes
  • .claude/settings.jsonasset707 bytes
  • .cursor/rules/workflow.mdcasset3,883 bytes
  • .cursor/skills/skill-based-architecture/SKILL.mddoc1,862 bytes
  • .gitignoreasset26 bytes
  • .maintenance-log.yamlasset1,703 bytes
  • AGENTS.mddoc3,900 bytes
  • CLAUDE.mddoc3,784 bytes
  • CODEX.mddoc4,091 bytes
  • EXAMPLES.mddoc7,795 bytes
  • GEMINI.mddoc3,784 bytes
  • LICENSEasset1,065 bytes
  • README.mddoc17,657 bytes
  • README.zh-CN.mddoc16,222 bytes
  • REFERENCE.mddoc7,226 bytes
  • SKILL.mdcore13,212 bytes
  • SKILL.mdcore13,212 bytes
  • SKILL.mdcore13,212 bytes
  • TEMPLATES-GUIDE.mddoc11,205 bytes
  • UPSTREAM-CHANGES-archive.mddoc222,512 bytes
  • UPSTREAM-CHANGES.mddoc18,990 bytes
  • WORKFLOW.mddoc12,569 bytes
  • assets/skill-based-architecture-title.pngasset3,411 bytes
  • docs/diagrams/README.mddoc7,451 bytes
  • docs/diagrams/file-call-graph-fixbug.pngasset151,607 bytes
  • docs/diagrams/file-call-graph-fixbug.svgasset6,084 bytes
  • docs/diagrams/progressive-rigor.pngasset88,160 bytes
  • docs/diagrams/progressive-rigor.svgasset3,872 bytes
  • docs/diagrams/routes-to-workflows.pngasset124,081 bytes
  • docs/diagrams/routes-to-workflows.svgasset5,119 bytes
  • docs/diagrams/routing-core.pngasset244,956 bytes
  • docs/diagrams/routing-core.svgasset8,419 bytes
  • docs/diagrams/skill-flow-complete.pngasset395,275 bytes
  • docs/diagrams/skill-flow-complete.svgasset13,987 bytes
  • docs/diagrams/system-panorama.pngasset260,186 bytes
  • docs/diagrams/system-panorama.svgasset9,988 bytes
  • docs/diagrams/upstream-downstream-sync.pngasset136,564 bytes
  • docs/diagrams/upstream-downstream-sync.svgasset6,024 bytes
  • docs/linuxdo-project-introduction.mddoc16,532 bytes
  • docs/plans/2026-05-12-thin-shells-generator.mddoc3,688 bytes
  • docs/plans/2026-05-20-subagent-surface-hints.mddoc19,809 bytes
  • docs/plans/2026-06-24-skeleton-flesh-axis/plan.mddoc13,143 bytes
  • docs/plans/2026-06-24-skeleton-flesh-axis/prd.mddoc5,698 bytes
  • docs/plans/2026-07-09-dynamic-rigor-assembly.mddoc13,707 bytes
  • docs/plans/2026-07-16-business-global-model-and-load-reason/prd.mddoc35,643 bytes
  • docs/plans/2026-07-27-rule-ownership-and-change-fanout/prd.mddoc21,922 bytes
  • docs/plans/2026-07-27-user-confirmed-plan-mainline/prd.mddoc7,200 bytes
  • docs/plans/2026-07-30-evidence-driven-progressive-disclosure/prd.mddoc29,191 bytes
  • docs/plans/2026-07-31-evidence-funnel-code-localization.mddoc34,224 bytes
  • docs/plans/2026-07-31-plan-feature-workflow-redesign/design.mddoc18,440 bytes
  • docs/plans/2026-07-31-plan-feature-workflow-redesign/prd.mddoc12,864 bytes
  • docs/plans/2026-08-03-plan-composition-and-db-artifacts/prd.mddoc43,188 bytes
  • docs/plans/2026-08-03-requirement-semantics-change-contract/prd.mddoc40,686 bytes
  • docs/plans/2026-08-04-default-downstream-simplification/prd.mddoc27,827 bytes
  • docs/plans/2026-08-04-failure-driven-long-term-memory.mddoc28,916 bytes
  • docs/plans/2026-08-04-proactive-workflow-distillation.mddoc26,541 bytes
  • docs/plans/2026-08-05-rule-update-progressive-disclosure/prd.mddoc30,957 bytes
  • docs/plans/2026-08-06-chaos-runtime-context-reduction/prd.mddoc75,130 bytes
  • docs/plans/2026-08-12-bounded-deterministic-execution.mddoc27,871 bytes
  • docs/plans/2026-08-12-requirement-first-planning.mddoc18,640 bytes
  • docs/plans/2026-08-13-requirement-definition-plan-separation/design.mddoc16,208 bytes
  • docs/plans/2026-08-13-requirement-definition-plan-separation/prd.mddoc10,336 bytes
  • docs/plans/README.mddoc15,324 bytes
  • docs/plans/_TEMPLATE.mddoc5,854 bytes
  • docs/sba-bible.mddoc21,227 bytes
  • docs/task-anchor-native-plan.mddoc17,073 bytes
  • examples/behavior-failures.mddoc16,803 bytes
  • examples/simple-repo/COPY-PASTE-INPUT.mddoc4,432 bytes
  • examples/simple-repo/EXPECTED-SHAPE.mddoc1,580 bytes
  • examples/simple-repo/README.mddoc2,329 bytes
  • examples/simple-repo/repo/.cursor/rules/frontend.mdcasset839 bytes
  • examples/simple-repo/repo/AGENTS.mddoc1,457 bytes
  • examples/simple-repo/repo/CLAUDE.mddoc920 bytes
  • examples/simple-repo/repo/README.mddoc549 bytes
  • examples/simple-repo/repo/package.jsonruns163 bytes
  • examples/simple-repo/repo/src/cart.jsruns1,032 bytes
  • references/README.mddoc515 bytes
  • references/business-global-model.mddoc18,580 bytes
  • references/conventions.mddoc11,508 bytes
  • references/executable-skill-architecture.mddoc7,755 bytes
  • references/layout.mddoc14,463 bytes
  • references/multi-skill-routing.mddoc13,713 bytes
  • references/per-tool-shells.mddoc6,876 bytes
  • references/permission-model.mddoc4,904 bytes
  • references/progressive-rigor.mddoc10,661 bytes
  • references/protocols.mddoc27,587 bytes
  • references/scenario-testing.mddoc6,388 bytes
  • references/self-hosting-conformance.yamlasset49,500 bytes
  • references/self-hosting-routing.yamlasset5,017 bytes
  • references/self-hosting-shell-base.mddoc2,930 bytes
  • references/self-hosting-shells.yamlasset2,927 bytes
  • references/skeleton-flesh-split.mddoc15,907 bytes
  • references/skill-composition.mddoc7,533 bytes
  • references/tests-as-spec.mddoc6,721 bytes
  • references/thin-shells.mddoc16,465 bytes
  • scripts/README.mddoc13,976 bytes
  • scripts/check-all.shruns39,246 bytes
  • scripts/check-live-agent-journey.shruns11,239 bytes
  • scripts/check-migration-evidence.shruns9,219 bytes
  • scripts/check-self-scenarios.shruns87,677 bytes
  • scripts/check-self-shells.shruns2,887 bytes
  • scripts/check-template-hooks.shruns4,382 bytes
  • scripts/check-upstream-changes.shruns3,998 bytes
  • scripts/check-upstream-supersedes.shruns2,618 bytes
  • scripts/check-validation-contract.shruns100,522 bytes
  • scripts/scaffold-downstream.shruns63,716 bytes
  • scripts/skill-assetruns19,433 bytes
  • scripts/sync-self-shells.shruns10,246 bytes
  • scripts/test.jsruns7,731 bytes
  • skill.yamlasset2,480 bytes
  • templates/ANTI-TEMPLATES.mddoc11,941 bytes
  • templates/README.mddoc11,384 bytes
  • templates/hooks/README.mddoc3,722 bytes
  • templates/hooks/SECURITY.mddoc2,080 bytes
  • templates/hooks/hooks-cursor.jsonasset904 bytes
  • templates/hooks/hooks.jsonasset921 bytes
  • templates/hooks/session-startruns2,378 bytes
  • templates/hooks/workflow-stateruns2,640 bytes
  • templates/shells/.cursor/rules/workflow.mdcasset3,820 bytes
  • templates/shells/.cursor/skills/{{NAME}}/SKILL.md.templateasset1,919 bytes
  • templates/shells/.cursor/skills/{{NAME}}/SKILL.single.md.templateasset513 bytes
  • templates/shells/AGENTS.mddoc4,244 bytes
  • templates/shells/CLAUDE.mddoc4,409 bytes
  • templates/shells/CODEX.mddoc4,230 bytes
  • templates/shells/GEMINI.mddoc4,231 bytes
  • templates/shells/single-entry.md.templateasset842 bytes
  • templates/skill/.upstream-syncasset472 bytes
  • templates/skill/SKILL.md.templateasset6,207 bytes
  • templates/skill/SKILL.routed.md.templateasset2,000 bytes
  • templates/skill/SKILL.single.md.templateasset2,244 bytes
  • templates/skill/conformance.yamlasset44,620 bytes
  • templates/skill/protocol-blocks/ambiguous-request-gate.mddoc3,213 bytes
  • templates/skill/protocol-blocks/change-contract.mddoc5,085 bytes
  • templates/skill/protocol-blocks/composite-plan-execution.mddoc2,480 bytes
  • templates/skill/protocol-blocks/external-delivery-verification.mddoc4,620 bytes
  • templates/skill/protocol-blocks/plan-knowledge-closure.mddoc3,390 bytes
  • templates/skill/protocol-blocks/reboot-check.mddoc1,958 bytes
  • templates/skill/protocol-blocks/source-localization.mddoc3,379 bytes
  • templates/skill/protocol-blocks/subagent-contract.mddoc2,572 bytes
  • templates/skill/protocol-blocks/user-decision-drift.mddoc2,881 bytes
  • templates/skill/protocol-blocks/workflow-distillation-proposal.mddoc2,021 bytes
  • templates/skill/references/behavior-failures.mddoc2,133 bytes
  • templates/skill/references/gotchas.mddoc1,150 bytes
  • templates/skill/references/subagent-verification.mddoc4,090 bytes
  • templates/skill/routing.yamlasset4,395 bytes
  • templates/skill/rules/change-discipline.mddoc1,687 bytes
  • templates/skill/rules/coding-standards.mddoc508 bytes
  • templates/skill/rules/project-rules.mddoc469 bytes
  • templates/skill/scripts/audit-orphans.shruns5,629 bytes
  • templates/skill/scripts/check-version-conformance.shruns15,109 bytes
  • templates/skill/scripts/route-health.shruns7,139 bytes
  • templates/skill/scripts/route-reachability.shruns7,380 bytes
  • templates/skill/scripts/smoke-test.shruns58,372 bytes
  • templates/skill/scripts/sync-routing.shruns31,244 bytes
  • templates/skill/scripts/sync-vendor.shruns5,749 bytes
  • templates/skill/scripts/upstream-status.shruns5,662 bytes
  • templates/skill/sync-manifest.yamlasset1,467 bytes
  • templates/skill/workflows/change-managed.mddoc5,252 bytes
  • templates/skill/workflows/define-requirement.mddoc7,653 bytes
  • templates/skill/workflows/edit-templates.mddoc3,800 bytes
  • templates/skill/workflows/fix-bug.mddoc9,302 bytes
  • templates/skill/workflows/invoke-skill.md.exampleasset5,416 bytes
  • templates/skill/workflows/light-task.md.templateasset872 bytes
  • templates/skill/workflows/maintain-docs.mddoc20,861 bytes
  • templates/skill/workflows/plan-feature.mddoc9,885 bytes
  • templates/skill/workflows/plan-large.mddoc5,216 bytes
  • templates/skill/workflows/profile-business-model.md.exampleasset14,549 bytes
  • templates/skill/workflows/profile-project.mddoc6,998 bytes
  • templates/skill/workflows/receiving-review.mddoc2,059 bytes
  • templates/skill/workflows/refactor-fanout.mddoc2,240 bytes
  • templates/skill/workflows/rule-update/business-truth.mddoc5,598 bytes
  • templates/skill/workflows/rule-update/knowledge-reconciliation.mddoc3,422 bytes
  • templates/skill/workflows/rule-update/verified-failure.mddoc5,921 bytes
  • templates/skill/workflows/rule-update/workflow-distillation.mddoc3,240 bytes
  • templates/skill/workflows/subagent-auxiliary.mddoc2,383 bytes
  • templates/skill/workflows/subagent-driven.mddoc2,295 bytes
  • templates/skill/workflows/subagent-orchestration.mddoc3,402 bytes
  • templates/skill/workflows/task-closure.materialized.md.templateasset722 bytes
  • templates/skill/workflows/task-closure.mddoc10,745 bytes
  • templates/skill/workflows/task-execution.materialized.md.templateasset2,046 bytes
  • templates/skill/workflows/task-execution.mddoc16,232 bytes
  • templates/skill/workflows/update-rules.mddoc7,827 bytes
  • templates/skill/workflows/update-upstream.mddoc10,090 bytes
  • workflows/full-migration.mddoc19,510 bytes
  • workflows/upgrade-downstream.mddoc5,470 bytes
Read SKILL.md

# Skill-Based Architecture

Restructure oversized single-file Skills or scattered project rules into a well-organized Skill directory. Builds on the official minimal Agent Skill contract (`name` + `description`) and kicks in when a single small `SKILL.md` is no longer enough.

## When to Use

- A single `SKILL.md` exceeds ~150 lines, mixing rules, workflows, and background material; or project rules are scattered across `AGENTS.md`, `CLAUDE.md`, `CODEX.md`, `.cursor/rules/`, `.claude/`, etc.
- Recurring tasks need procedures, fitted verification, or reliable completion checks; or the user explicitly requests Skill-based architecture or rule consolidation.
- Skip temporary repos with no durable routing need, teams with a working compact instruction system, and small projects; an explicit SBA request may still materialize a complete direct `SKILL.md` carrier.

## Progressive Rigor

Grow only under pressure. The materializer derives **Single-file**, **Folder-light**, or **Full/broad** internally from target evidence; users do not select a tier, profile, capability pack, or install mode. `routing.yaml`, Task Execution, Task Closure, maintenance checks, and each harness surface need their own routing/loading/ownership pressure. **Split by abstraction (骨架/肉)** when content tangles invariant design theory with current-code facts: abstract theory → `architecture/`, code maps → `references/`, house style → `conventions/`, per-module landmines → `gotchas/` (methodology stays in `rules/`). Downgrade when content shrinks. Details: [references/progressive-rigor.md](references/progressive-rigor.md).
View the complete source on GitHub

Skills included

  • skill-based-architecture

Repository facts

License
MIT
Created
Apr 8, 2026
Last pushed
Sep 20, 2026
Forks
52
Open issues
0
Project activity
Active

Before you install

Network or broad shell

Automated signals, not a guarantee

Checked Nov 4, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 2 notes
  • SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 183 signals by file

.claude/hooks/session-start

  • Executable scriptline 1
    #!/usr/bin/env bash

README.md

  • Network accessline 20
    <p align="center">   <a href="https://github.com/WoJiSama/skill-based-architecture/stargazers">
  • Network accessline 21
    <a href="https://github.com/WoJiSama/skill-based-architecture/stargazers">
  • Network accessline 22
    <img alt="GitHub stars" src="https://img.shields.io/github/stars/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 23
    </a>   <a href="https://github.com/WoJiSama/skill-based-architecture/forks">     <img alt="GitHub forks" src="https://img.shields.io/github/forks/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 24
    <a href="https://github.com/WoJiSama/skill-based-architecture/forks">
  • Network accessline 25
    <img alt="GitHub forks" src="https://img.shields.io/github/forks/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 26
    </a>   <a href="LICENSE">     <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">
  • Network accessline 27
    <a href="LICENSE">     <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">   </a>
  • Network accessline 28
    <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">
  • Network accessline 79
    git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 111
    - Use [`examples/simple-repo/`](examples/simple-repo/) as a tiny local fixture with deliberately duplicated instructions. - Use [`COPY-PASTE-INPUT.md`](examples/simple-repo/COPY-PASTE-INPUT.md) when a hosted environment accepts pasted files
  • Network accessline 113
    - [Run SBA in Telegram or WhatsApp through ClawMama](https://app.clawmama.run/skills/i78bb1/hermes?utm_source=github&utm_medium=issue&utm_campaign=skill_outreach_wojisama_skill_based_architecture) for a lightweight external preview. Do not 
  • Network accessline 245
    License: [LICENSE](LICENSE). Community discussion: [LinuxDO](https://linux.do/).
  • Network accessline 249
    <a href="https://www.star-history.com/?repos=WoJiSama%2Fskill-based-architecture&type=date&legend=top-left">
  • Network accessline 250
    [Credential-related source omitted]
  • Network accessline 251
    [Credential-related source omitted]
  • Network accessline 252
    [Credential-related source omitted]
  • Network accessline 253
    [Credential-related source omitted]

README.zh-CN.md

  • Network accessline 20
    <p align="center">   <a href="https://github.com/WoJiSama/skill-based-architecture/stargazers">
  • Network accessline 21
    <a href="https://github.com/WoJiSama/skill-based-architecture/stargazers">
  • Network accessline 22
    <img alt="GitHub stars" src="https://img.shields.io/github/stars/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 23
    </a>   <a href="https://github.com/WoJiSama/skill-based-architecture/forks">     <img alt="GitHub forks" src="https://img.shields.io/github/forks/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 24
    <a href="https://github.com/WoJiSama/skill-based-architecture/forks">
  • Network accessline 25
    <img alt="GitHub forks" src="https://img.shields.io/github/forks/WoJiSama/skill-based-architecture?style=flat&logo=github">
  • Network accessline 26
    </a>   <a href="LICENSE">     <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">
  • Network accessline 27
    <a href="LICENSE">     <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">   </a>
  • Network accessline 28
    <img alt="License" src="https://img.shields.io/github/license/WoJiSama/skill-based-architecture?style=flat">
  • Network accessline 79
    git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 111
    - 把 [`examples/simple-repo/`](examples/simple-repo/) 当作本地最小 fixture;其中故意放了互相重复的指令。 - Hosted 环境支持粘贴文件时,使用 [`COPY-PASTE-INPUT.md`](examples/simple-repo/COPY-PASTE-INPUT.md)。 - 通过 ClawMama [在 Telegram 或 WhatsApp 中运行 SBA](https://app.clawmama.r
  • Network accessline 113
    - 通过 ClawMama [在 Telegram 或 WhatsApp 中运行 SBA](https://app.clawmama.run/skills/i78bb1/hermes?utm_source=github&utm_medium=issue&utm_campaign=skill_outreach_wojisama_skill_based_architecture),进行轻量外部体验。不要上传敏感项目规则。
  • Network accessline 245
    许可证:[LICENSE](LICENSE)。社区讨论:[LinuxDO](https://linux.do/)。
  • Network accessline 249
    <a href="https://www.star-history.com/?repos=WoJiSama%2Fskill-based-architecture&type=date&legend=top-left">
  • Network accessline 250
    [Credential-related source omitted]
  • Network accessline 251
    [Credential-related source omitted]
  • Network accessline 252
    [Credential-related source omitted]
  • Network accessline 253
    [Credential-related source omitted]

WORKFLOW.md

  • Dangerous shell operationline 118
    trap 'rm -f "$MIGRATION_EVIDENCE_JSON"' EXIT

docs/diagrams/file-call-graph-fixbug.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="620" viewBox="0 0 680 620" role="img">

docs/diagrams/progressive-rigor.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="330" viewBox="0 0 680 330" role="img">

docs/diagrams/routes-to-workflows.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="540" viewBox="0 0 680 540" role="img">

docs/diagrams/routing-core.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="800" viewBox="0 0 680 800" role="img">

docs/diagrams/skill-flow-complete.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="1420" viewBox="0 0 680 1420" role="img">

docs/diagrams/system-panorama.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="906" viewBox="0 0 680 906" role="img">

docs/diagrams/upstream-downstream-sync.svg

  • Network accessline 1
    <svg xmlns="http://www.w3.org/2000/svg" width="680" height="502" viewBox="0 0 680 502" role="img">

docs/linuxdo-project-introduction.md

  • Network accessline 22
    项目地址:[WoJiSama/skill-based-architecture](https://github.com/WoJiSama/skill-based-architecture)
  • Network accessline 90
    如果一个 skill 只有几条稳定原则,单个 `SKILL.md` 就够了。参考 [forrestchang/andrej-karpathy-skills](https://github.com/forrestchang/andrej-karpathy-skills) 这类项目,核心价值不是目录复杂,而是规则写得克制、具体、可检查。
  • Network accessline 395
    # Cursor 用户级 skill git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 396
    git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 399
    # Cursor 项目级 skill git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 400
    git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 403
    # 通用项目内安装 git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 404
    git clone https://github.com/WoJiSama/skill-based-architecture.git \
  • Network accessline 455
    [https://github.com/WoJiSama/skill-based-architecture](https://github.com/WoJiSama/skill-based-architecture)
  • Network accessline 461
    <a href="https://www.star-history.com/?repos=WoJiSama%2Fskill-based-architecture&type=date&legend=top-left">
  • Network accessline 462
    <picture>    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=WoJiSama/skill-based-architecture&type=date&theme=dark&legend=top-left" />    <source media="(prefers-color-scheme: light)" srcset="h
  • Network accessline 463
    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=WoJiSama/skill-based-architecture&type=date&theme=dark&legend=top-left" />
  • Network accessline 464
    <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=WoJiSama/skill-based-architecture&type=date&legend=top-left" />
  • Network accessline 465
    <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=WoJiSama/skill-based-architecture&type=date&legend=top-left" />

docs/plans/2026-07-27-rule-ownership-and-change-fanout/prd.md

  • Network accessline 217
    - Upstream ownership convergence landed in commit `5a9785d`. The complete Drift Gate is owned by `task-execution.md`; Plan, information-model, distillation, documentation-maintenance, and Closure responsibilities remain with their own route
  • Network accessline 218
    - Upstream closure passed `scripts/check-all.sh`, including both temporary downstream journeys, self-hosting scenarios, template/self-hosting conformance, shell/path integrity, and orphan/reachability checks. The D7 simulation found one com
  • Network accessline 219
    - Chaos adaptation landed in exact meta commit `5bad9fd` and was pushed on `codex/rule-ownership-local-action-20260727`. Codeup MR [#17](https://codeup.aliyun.com/60dd34bf4690c27532d3d021/aiinfra/chaos/change/17) is open and pending from th

docs/plans/2026-08-03-plan-composition-and-db-artifacts/prd.md

  • Network accessline 547
    - **0.77 assembly:** `aii update release-0.77_shiqi --no-pull --no-self-upgrade` assembled the canonical commit. Backend and frontend canonical, `.codex`, and `.claude` trees are recursively identical, while both business repositories retai
  • Network accessline 548
    - **Database-change dossier:** the 0.77 iteration-testing-mode `prd.md` now has a separate schema-impact checkpoint and its sibling `schema.sql` contains the live MySQL 8.0.27 baseline, complete target DDL, forward schema/data update, read-
  • Network accessline 549
    - **Canonical MR:** Codeup MR `#22` is `https://codeup.aliyun.com/60dd34bf4690c27532d3d021/aiinfra/chaos/change/22`, source `codex/plan-composition-db-artifacts-20260803`, target `master`, state `待合并`, one commit, eight files, and `无代码冲突`. 

docs/plans/2026-08-04-failure-driven-long-term-memory.md

  • Instruction override signalline 191
    - Consumes: T1-T3 final wording. - Produces: owner/hook assertions and transcript-shaped failure-learning scenarios. - Acceptance: checks fail if Rule Update loses complete ownership, consumers copy its procedure, first failures create a le
  • Instruction override signalline 193
    - Acceptance: checks fail if Rule Update loses complete ownership, consumers copy its procedure, first failures create a ledger, retries count as recurrence, activation repair is skipped, or shared gates bypass approval.

examples/simple-repo/README.md

  • Network accessline 30
    Use the files under: https://github.com/WoJiSama/skill-based-architecture/tree/main/examples/simple-repo/repo
  • Network accessline 31
    https://github.com/WoJiSama/skill-based-architecture/tree/main/examples/simple-repo/repo

examples/simple-repo/repo/package.json

  • Executable scriptline 1
    Executable file

examples/simple-repo/repo/src/cart.js

  • Executable scriptline 1
    Executable file

references/layout.md

  • Network accessline 173
    Anthropic's public [`skills` repository](https://github.com/anthropics/skills) defines the **minimal** skill shape: a folder with a `SKILL.md`, plus frontmatter where `name` identifies the skill and `description` explains what it does and w

references/per-tool-shells.md

  • Network accessline 97
    <!-- external-fact: verified=2026-04-28 source=https://code.claude.com/docs/en/skills -->
  • Network accessline 121
    <!-- external-fact: verified=2026-04-28 source=https://docs.cursor.com/en/context -->
  • Network accessline 122
    <!-- external-fact: verified=2026-04-28 source=https://code.claude.com/docs/en/skills -->
  • Network accessline 123
    <!-- external-fact: verified=2026-04-28 source=https://developers.openai.com/codex/guides/agents-md -->
  • Network accessline 124
    <!-- external-fact: verified=2026-04-28 source=https://docs.windsurf.com/windsurf/cascade/memories -->
  • Network accessline 125
    <!-- external-fact: verified=2026-04-28 source=https://github.com/google-gemini/gemini-cli/blob/main/docs/cli/gemini-md.md -->
  • Network accessline 126
    <!-- external-fact: verified=2026-04-28 source=https://opencode.ai/docs/rules/ -->

references/skill-composition.md

  • Network accessline 7
    A project skill captures *this project's* rules and routing. General-purpose capabilities — planning, security review, writing tests, research — are better built once in a reusable skill (e.g. [`obra/superpowers`](https://github.com/obra/su

references/thin-shells.md

  • Network accessline 119
    The pattern is adopted from [OpenSpec](https://github.com/Fission-AI/OpenSpec)'s `<context>` / `<rules>` injection approach, adapted to our routing-table model.

scripts/check-all.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 116
    find . -name '*.bak' -type f -delete
  • Dangerous shell operationline 193
    trap 'rm -rf "$tmp" "$before"' EXIT
  • Dangerous shell operationline 241
    trap 'rm -rf "$tmp" "$before"' EXIT
  • Dangerous shell operationline 322
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 361
    trap 'rm -rf "$tmp" "$before" "$after_apply"' EXIT
  • Dangerous shell operationline 535
    rm -rf "$tmp"
  • Dangerous shell operationline 540
    rm -rf "$tmp"
  • Dangerous shell operationline 544
    rm "$skill_root/gotchas/shared.md"
  • Dangerous shell operationline 549
    rm -rf "$tmp"
  • Dangerous shell operationline 574
    rm -rf "$tmp"
  • Dangerous shell operationline 577
    rm -rf "$tmp"
  • Dangerous shell operationline 584
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 632
    rm -rf "$skill/scripts"
  • Dangerous shell operationline 643
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 736
    rm "$skill/references/business/unregistered.md"
  • Dangerous shell operationline 760
    rm "$skill/references/navigation.md"
  • Dangerous shell operationline 775
    rm "$root/services/billing/skills/billing/references/business/billing.md"
  • Dangerous shell operationline 807
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 888
    trap 'rm -rf "$tmp"' EXIT

scripts/check-live-agent-journey.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 62
    trap 'rm -rf "$tmp"' EXIT

scripts/check-migration-evidence.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/check-self-scenarios.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/check-self-shells.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/check-template-hooks.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Access to secretsline 40
    [Credential-related source omitted]
  • Dangerous shell operationline 94
    trap 'rm -rf "$TMP"' EXIT

scripts/check-upstream-changes.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/check-upstream-supersedes.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/check-validation-contract.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 29
    find "$target/skills/$name" -maxdepth 1 -type f      -name 'SKILL.*.md.template' ! -name 'SKILL.md.template' -delete
  • Dangerous shell operationline 35
    find "$target/skills/$name" -type f \( -name '*.md.template' -o -name '*.yaml.template' \) -delete
  • Dangerous shell operationline 60
    find . -name '*.bak' -type f -delete
  • Dangerous shell operationline 188
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 238
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 417
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 498
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 524
    find . -name '*.bak' -type f -delete
  • Dangerous shell operationline 550
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 774
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 805
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 855
    rm -f "$owner.bak"
  • Dangerous shell operationline 974
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1014
    rm -f "$owner.bak"
  • Dangerous shell operationline 1040
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1495
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1620
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1728
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1850
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1911
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 1997
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2061
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2130
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2166
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2265
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2309
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2339
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2369
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2411
    trap 'rm -rf "$tmp"' EXIT
  • Dangerous shell operationline 2433
    trap 'rm -rf "$tmp"' EXIT

scripts/scaffold-downstream.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 1424
    find "$stage" -name '*.bak' -type f -delete
  • Dangerous shell operationline 1537
    cleanup() { rm -rf "$stage"; }
  • Dangerous shell operationline 1543
    for path in "${created_paths[@]}"; do rm -rf "$TARGET/$path"; done

scripts/skill-asset

  • Executable scriptline 1
    #!/usr/bin/env node

scripts/sync-self-shells.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

scripts/test.js

  • Executable scriptline 1
    #!/usr/bin/env node
  • Possible obfuscationline 212
    fs.writeFileSync(path.join(rulesDir, 'bom.md'), '<U+FEFF>## BomHeadingZzz\n\nContent about BomKeyword99.\n');

templates/hooks/README.md

  • Network accessline 59
    Schema check: Claude Code CLI v2.1+ uses the nested hook format (`matcher` -> `hooks[]` -> `{type,command}`). If a hook appears configured but does not fire, inspect hook events with the CLI's verbose hook output. <!-- external-fact: verifi
  • Network accessline 60
    <!-- external-fact: verified=2026-05-06 source=https://code.claude.com/docs/en/hooks -->

templates/hooks/SECURITY.md

  • Network accessline 19
    Origin: trust-boundary principle condensed from [OthmanAdi/planning-with-files](https://github.com/OthmanAdi/planning-with-files).

templates/hooks/session-start

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/hooks/workflow-state

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/shells/CLAUDE.md

  • Network accessline 5
    <!-- external-fact: verified=2026-04-28 source=https://code.claude.com/docs/en/skills -->

templates/skill/.upstream-sync

  • Network accessline 3
    # - workflows/update-upstream.md sets synced_sha (final step) after each refresh. # - scripts/upstream-status.sh reads it to report what's new upstream since then. upstream: https://github.com/WoJiSama/skill-based-architecture
  • Network accessline 4
    # - scripts/upstream-status.sh reads it to report what's new upstream since then. upstream: https://github.com/WoJiSama/skill-based-architecture synced_sha: <set by update-upstream.md on first refresh>
  • Network accessline 5
    upstream: https://github.com/WoJiSama/skill-based-architecture

templates/skill/protocol-blocks/reboot-check.md

  • Network accessline 21
    Origin: condensed from [OthmanAdi/planning-with-files](https://github.com/OthmanAdi/planning-with-files) 5-Question Reboot Test.

templates/skill/scripts/audit-orphans.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/skill/scripts/check-version-conformance.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 53
    trap 'rm -f "$PARSED_FILE"' EXIT

templates/skill/scripts/route-health.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/skill/scripts/route-reachability.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/skill/scripts/smoke-test.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 1250
    rm -f "$REFERENCE_REPORT" "$REFERENCE_INPUT"

templates/skill/scripts/sync-routing.sh

  • Executable scriptline 1
    #!/usr/bin/env bash

templates/skill/scripts/sync-vendor.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 50
    TMP_DIR="$(mktemp -d)"; trap 'rm -rf "$TMP_DIR"' EXIT

templates/skill/scripts/upstream-status.sh

  • Executable scriptline 1
    #!/usr/bin/env bash
  • Dangerous shell operationline 84
    CLEAN=""; trap '[[ -n "$CLEAN" ]] && rm -rf "$CLEAN"' EXIT

templates/skill/workflows/invoke-skill.md.example

  • Network accessline 19
    3. `rules/change-discipline.md` immediately before the first mutation 4. **Verify the invoked skill is present** — `ls skills/superpowers/SKILL.md` must exist. If not:    - Stop and report: "This workflow depends on skills/superpowers/. Ins
  • Network accessline 20
    4. **Verify the invoked skill is present** — `ls skills/superpowers/SKILL.md` must exist. If not:    - Stop and report: "This workflow depends on skills/superpowers/. Install it first: `git clone https://github.com/obra/superpowers skills/s
  • Network accessline 21
    - Stop and report: "This workflow depends on skills/superpowers/. Install it first: `git clone https://github.com/obra/superpowers skills/superpowers`."

templates/skill/workflows/update-upstream.md

  • Network accessline 5
    <!-- upstream-source: repo=https://github.com/WoJiSama/skill-based-architecture.git -->
  • Network accessline 32
    tmp="$(mktemp -d)"    git clone https://github.com/WoJiSama/skill-based-architecture.git "$tmp/upstream"
  • Network accessline 33
    git clone https://github.com/WoJiSama/skill-based-architecture.git "$tmp/upstream"
  • Network accessline 65
    ```bash     printf 'upstream: %s\nsynced_sha: %s\nsynced_date: %s\n' \       "https://github.com/WoJiSama/skill-based-architecture" \
  • Network accessline 66
    printf 'upstream: %s\nsynced_sha: %s\nsynced_date: %s\n' \       "https://github.com/WoJiSama/skill-based-architecture" \       "$(git -C "$tmp/upstream" rev-parse HEAD)" "$(date +%F)" > "skills/$NAME/.upstream-sync"
  • Network accessline 67
    "https://github.com/WoJiSama/skill-based-architecture" \

workflows/full-migration.md

  • Network accessline 199
    <!-- external-fact: verified=2026-04-28 source=https://code.claude.com/docs/en/skills -->

workflows/upgrade-downstream.md

  • Network accessline 10
    https://github.com/WoJiSama/skill-based-architecture.git

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Signal confidence: 170 facts

and 120 more

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativeManifest .claude-plugin/marketplace.json.claude/skills/<skill-name>/
CodexNativeManifest AGENTS.md.agents/skills/<skill-name>/
CursorNativePath .cursor/rules/workflow.mdc.cursor/skills/<skill-name>/
Gemini CLINativePath GEMINI.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.