ClawWork
AGENTS.md533 starsChanged 5 months agoApache-2.0View on GitHub
This is the canonical agent entrypoint for this repo. Tool-specific shims such as CLAUDE.md and .github/copilot-instructions.md must point here instead of duplicating project rules. If the task touches Gateway behavior, also inspect ~/git/openclaw before changing local code. Before claiming done: Project-specific skills may remain under .claude/skills/ for current tool discovery compatibility. Do not add generic or personal skills here; use ~/.agents/skills for shared local skills.
- Installs packages
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 32,671 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository7 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- pr-cleaner-masterSkill
- pr-merge-runnerSkill
- release-publishSkill
- sync-gateway-docsSkill
- team-createSkill
The file
We split this AGENTS.md into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 33
If the task touches Gateway behavior, also inspect ~/git/openclaw before changing local code.
- Distinctive: line 48
Do not commit local model preferences, personal approval rules, hook settings, API endpoints, tokens, workspace paths, or one-off planning …
≈ “Common” is an estimate: at least 1 in 5 of the 41,581 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
ClawWork Agent Guide
This is the canonical agent entrypoint for this repo. Tool-specific shims such
as CLAUDE.md and .github/copilot-instructions.md must point here instead of
duplicating project rules.
Quick Reference
- Product: OpenClaw desktop client (Electron 34 + React 19)
- Monorepo:
packages/shared/(types + protocol) ->packages/core/(stores, services, ports) ->packages/desktop/(Electron app) +packages/pwa/(web app) +website/+keynote/ - Gateway: single WS to
:18789, session keyagent:main:clawwork:task:<taskId> - Design doc:
docs/openclaw-desktop-design.md
Commands
pnpm install # dependencies
pnpm dev # desktop dev with hot reload
pnpm check # full verification gateRequired Reading Order
AGENTS.md— project overview, commands, verification gatedocs/agent-rules/architecture.md— product identity, layer ownership, invariantsdocs/agent-rules/frontend.md— TypeScript, React, stylingdocs/agent-rules/main-process.md— Electron main process, IPC, WebSocketdocs/agent-rules/message-persistence.md— message write paths (CRITICAL)docs/agent-rules/git-conventions.md— commit format, PR budgetdocs/openclaw-desktop-design.md— historical design context- The module you will change
If the task touches Gateway behavior, also inspect ~/git/openclaw before
changing local code.
Verification
Before claiming done:
pnpm checkpasses, or the remaining failure is explicitly scoped and reported- Message-related changes: SQLite duplicate query returns empty
- UI changes: verify in both dark and light themes
Local Overrides
- Personal preferences and machine-specific workflow belong in
AGENTS.local.md. - Claude-only preferences belong in
CLAUDE.local.mdor.claude/settings.local.json. - Do not commit local model preferences, personal approval rules, hook settings, API endpoints, tokens, workspace paths, or one-off planning notes.
Project Skills
Project-specific skills may remain under .claude/skills/ for current tool
discovery compatibility. Do not add generic or personal skills here; use
~/.agents/skills for shared local skills.
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.

