turbo-start-web-agent
CLAUDE.md4 starsChanged 47 days agoMITView on GitHub
Everything above this line lives in AGENTS.md, pulled in verbatim by the @AGENTS.md import above — Claude Code does not read AGENTS.md on its own, so this file is the only reason those rules are visible to it at all. The lines below apply only to Claude Code and have no equivalent in the platform-neutral file. - Cloud sessions clone this repo into an isolated VM and load .claude/settings.json hooks, .mcp.json, .claude/rules/, .claude/skills/, .claude/agents/, and .claude/commands/ from that clone —…
In this repository15 other files it gives its agents.
- AGENTS.mdAGENTS.md
- compose-pageSkill
- design-a-blockSkill
- find-skillsSkill
- humanizerSkill
- no-em-dash-copySkill
- sync-changesSkill
- write-a-skillSkill
- compose-pageSkill
- design-a-blockSkill
- find-skillsSkill
- humanizerSkill
- no-em-dash-copySkill
- sync-changesSkill
- write-a-skillSkill
The file
We split this CLAUDE.md into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are 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.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 24
Commit anything a cloud session needs to see.
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.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
<!--
GENERATED FILE — do not edit directly.
Source: AGENTS.md (edit the rules themselves) and
scripts/harness-gen/adapters/claude-code.ts (edit the Claude-specific
lines below). Regenerate with `pnpm harness`.
-->CLAUDE.md — Harbour
@AGENTS.md
Claude Code specifics
Everything above this line lives in AGENTS.md, pulled in verbatim by the
@AGENTS.md import above — Claude Code does not read AGENTS.md on its
own, so this file is the only reason those rules are visible to it at all.
The lines below apply only to Claude Code and have no equivalent in the
platform-neutral file.
- Cloud sessions clone this repo into an isolated VM and load
.claude/settings.jsonhooks,.mcp.json,.claude/rules/,.claude/skills/,.claude/agents/, and.claude/commands/from that clone — nothing under~/.claude/on anyone's own machine carries over. Commit anything a cloud session needs to see. - A
SessionStarthook in.claude/settings.jsonrunsscripts/preflight.sh, gated onCLAUDE_CODE_REMOTE=trueso it only fires in cloud sessions, and its report (git remote reachability,ghPR ability,pnpmusability, plus every capability declared inharness.config.json) is added to context before you make any claim about what you can push, publish, or open a PR for. Locally, runsh scripts/preflight.shby hand for the same report. .claude/skills/is a generated, byte-identical copy of.agents/skills/(see.claude/skills/README.md), not a symlink. Edit skills under.agents/skills/and runpnpm harnessto refresh the copy — the copy itself is generated output, exactly like this file.- Session-start git discipline (pull before editing, branch → commit → PR,
never push to
main, and how to tell whether a push actually reached the remote on platforms where the shell has no git credentials) lives in.agents/skills/sync-changes/SKILL.md— read it before your first edit in a fresh session.
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.

