claude-code-log
daaain/claude-code-log/CLAUDE.md
facing docs live in [docs/](docs/); plans and TODOs live in [work/](work/). A documentation site (MkDocs + Material) is published to GitHub Pages from `docs/` + `mkdocs.yml`. The CLI reference
How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
daaain/claude-code-log/CLAUDE.md
facing docs live in [docs/](docs/); plans and TODOs live in [work/](work/). A documentation site (MkDocs + Material) is published to GitHub Pages from `docs/` + `mkdocs.yml`. The CLI reference
modu-ai/moai-adk/CLAUDE.md
settings + template) — see §15. `MODE_TEAM_UNAVAILABLE` survives only as the historical fallback sentinel documented in `run.md`. Agent authoring: `.claude/rules/moai/development/agent-authoring.md`. --- ## 5. SPEC-Based Workflow MoAI uses
huytieu/COG-second-brain/CLAUDE.md
build-verify-fix pipeline is in `.claude/skills/closed-loop/SKILL.md`; multi-session goals are in `.claude/skills/ultragoal/SKILL.md`. Those documents oblige nothing in a session that never invoked them. Two of the harness's rules
microsoft/skills-for-fabric/CLAUDE.md
saved MCP configuration. ## Fabric REST APIs All Fabric operations use the REST APIs documented at: https://learn.microsoft.com/en-us/rest/api/fabric/articles/ ## Developer vs Consumer Patterns ### Developers - Use **REST APIs** to create/manage artifacts (workspaces
Chorus-AIDLC/Chorus/CLAUDE.md
with fine-grained, configurable permissions) and humans work together through a shared Idea → Proposal → Document + Task → Execute → Verify → Done pipeline. Core philosophy: **"Reversed Conversation"** — AI proposes, humans verify (not human
sheshbabu/zen/CLAUDE.md
install github.com/evanw/esbuild/cmd/esbuild@latest ``` ## Progressive disclosure — load only what you need Do not read every documentation file at the start of a task. - **Service boundaries, feature layout, routing, env vars, background
mattprusak/autoresearch-genealogy/CLAUDE.md
born`, `died`, `life_status`, `family`, `evidence_tier`, `profile_status`, `sources` - Transcription files add: `source`, `document_type`, `person`, `date`, `ocr_method`, `ocr_quality` - Region and surname files add: `evidence_tier` - Wikilinks
fcakyon/claude-codex-settings/.claude/CLAUDE.md
README) unless asked. - Prefer `rg` over `grep`. - Use the `writing-guidelines` skill to review documentation or interface copy, and the `humanize` skill to remove generic or machine-written phrasing
vm0-ai/vm0/CLAUDE.md
English. Use the user's preferred language in direct conversation. ## Task Routing Use [the documentation index](docs/docs.md) to select relevant guidance. | Changed surface | Read | | ------------------------------------------ | ----------------------------------------------------------------------------------------------------------- | | Tests or test failures | [Testing](docs/testing.md
RchGrav/claudebox/claude/CLAUDE.md
with behavioral changes - Any multi-step implementation work 4. **Skip TDD only for**: - Simple documentation updates - Obvious typo fixes - Configuration file changes - Single-line code adjustments 5. **TDD workflow activation
RchGrav/claudebox/CLAUDE.md
functions and clearly scoped variables are easier to test and reason about. | | **Lint, test, document** | Static analysis + automated tests + inline docs prevent regressions and knowledge rot. | --- ## 2 Mandatory Safety Flags
withkynam/vibecode-pro-max-kit/CLAUDE.md
PLAN/VALIDATE write artifacts with no implementation; EXECUTE implements the approved plan only; UPDATE PROCESS documents and archives. **Safety** - Never skip directly to implementation for substantial work - Never modify files
weiesky/cc-viewer/CLAUDE.md
include a changeset file. Never hand-edit `pnpm-lock.yaml`. Code comments, commit messages, and inline documentation must be in English — no language mixing in the codebase. Swallowed catch blocks with diagnostic
asklokesh/loki-mode/CLAUDE.md
with `loki_run_tmp_cleanup`. Never sweep `/tmp`, `$TMPDIR`, or process names/patterns. Verify-before-documenting rule and full source commentary: `docs/dev/tmp-cleanup.md`. ```bash loki_run_tmp_create() { local temp_root marker
jarrodwatts/claude-code-config/CLAUDE.md
does our code work? | How does this library work? | | Project-specific logic | Official API documentation | | | Library best practices & quirks | | | OSS implementation examples | **Trigger phrases** (fire librarian immediately
vstorm-co/pydantic-deepagents/CLAUDE.md
based terminal AI assistant) - `apps/deepresearch/` — Full-featured research reference app - `tests/` — Unit tests - `docs/` — Documentation source (MkDocs) ### Core Components **Agent Factory (`pydantic_deep/agent.py`)** - `create_deep_agent()`: Main factory function
Mininglamp-OSS/octo-server/CLAUDE.md
blank import in `internal/modules.go` - Auth: all routes go through `AuthMiddleware` unless explicitly excluded — document why if skipping - i18n: user-facing errors use `httperr.ResponseErrorL` + a registered `pkg/errcode` code; never raw `c.ResponseError
nateherkai/hyperframes-student-kit/CLAUDE.md
natural language. Model and permission choices belong to the user; `.codex/config.toml` only adds document loading defaults. Work locally unless the user requests subagents. ## Start and route Read `README.md` for installation
jonwiggins/optio/CLAUDE.md
legacy, local dev only): reads from host macOS Keychain ### Key subsystems These are well-documented in code; read the relevant service files for details: - **PR watcher** (`pr-watcher-worker.ts`): polls PRs every
yuanchenglu/claude-code-webui/CLAUDE.md
dialogs, etc.) │ │ ├── types/ # Type definitions │ │ └── contexts/ # React contexts ├── shared/ # Shared TypeScript types └── CLAUDE.md # Technical documentation ``` ## Key Design Decisions 1. **Runtime Abstraction**: Platform-agnostic business logic with minimal Runtime interface
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.