naman10parikhCLAUDE.mdagentgrid
naman10parikh/agentgrid/CLAUDE.md
Forged from Energy via harness-forge (CP103). One repo = one recursively self-improving agent-native harness. Energy is the control center; this is a self-contained flavor.
CLAUDE.md examples from real projects
How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
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.
Worked for most · soon
naman10parikhCLAUDE.mdnaman10parikhReads credentialsCLAUDE.mdagentcast
naman10parikh/agentcast/CLAUDE.md
Claude Code resolves project sub-agent personas from .claude/agents/<name>.md. NEVER create agents/agents/<name>.md — this is the A3 anti-pattern. The validator hard-fails on it. Enforce at all times. - Frontmatter: name and description are required; add a minimal tools list (the key is tools; Claude Code ignores allowed-tools in a sub-agent) - Model: leave model out, so the persona runs on the session's model; set it only when a role truly needs to - Read-only personas: include "DO NOT MODIFY FILES" in…
naman10parikhCLAUDE.mdfleet-audit
naman10parikh/fleet-audit/CLAUDE.md
- The hook. .claude/settings.json runs hooks/prepushscrubhook.py before every Bash command. On a git push it scrubs the committed tree with the patterns file in $FLEETAUDITPATTERNS (or .fleet-audit/scrub-patterns.txt, which git ignores) and blocks the push on a match, naming the pattern number and the file. With no patterns file it lets the push through. The plugin ships the same hook in hooks/hooks.json. - The MCP server. .mcp.json runs the server from GitHub through uvx. To try your local changes, start Claude…
naman10parikhReads credentialsCLAUDE.mdpixel
naman10parikh/pixel/CLAUDE.md
One repo = one recursively self-improving agent-native harness. Energy is the control center; pixel is the frontend (web) employee of the fleet. Read brain/ first — the entry point is brain/MOC - pixel.md. Identity: identity/SOUL.md. Persona/colors: identity/BRAND.md. Bootstrap facts: identity/MEMORY.md. AGENTS.md (imported above) says what pixel does, its tools and what it refuses. This file adds what Claude Code needs to work inside this repository: the pins, the loop, the hooks and the MCP setup. Next.js 16.2.9 · React 19.2.7…
naman10parikhReads credentialsCLAUDE.mdturbo
naman10parikh/turbo/CLAUDE.md
The Energy fleet's backend engineer. A coding agent's senior backend/runtime expert — consumed over CLI + MCP by agents, watched through a local cockpit by the chairman. Powered by Energy. Brain = brain/ (six Wave-R artifacts). Read identity/SOUL.md for the full persona, this file for how to operate. (Inherited from energy via the fleet scaffold; rewritten real for turbo — CP123.) A domain-expert employee for TS-strict backend services on Node 24 / Cloudflare Workers, framework default Hono. turbo scaffolds and…
About CLAUDE.md
What is CLAUDE.md?
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
Where does it go?
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
What should it contain?
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.md or AGENTS.md?
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.