harness-kit / templates
CLAUDE.md2 starsChanged 28 days agoMITView on GitHub
The import above is the whole hookup: Claude Code loads CLAUDE.md, which pulls in the engine-neutral canon. Two things to know: 1. An @import does not save context. Imported files load at session start, so CLAUDE.md + AGENTS.md cost exactly what one merged file would. The split exists only so other engines (which read AGENTS.md) see the same rules. If Claude Code is your only engine, you may rename AGENTS.md to CLAUDE.md and delete this file - one file is…
In this repository6 other files it gives its agents.
- templates/AGENTS.mdAGENTS.md
- absence-checkSkill
- canon-upkeepSkill
- harness-setupSkill
- independent-reviewSkill
- pattern-cardSkill
The file
We split this CLAUDE.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 15
Do not list your skills or subagents here.
- Distinctive: line 26
Add below only rules that are true for Claude Code and false for every other engine.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 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
@AGENTS.md
<!--
harness-kit template v0.2 - Claude Code appendix.
The import above is the whole hookup: Claude Code loads CLAUDE.md, which
pulls in the engine-neutral canon. Two things to know:
1. An @import does not save context. Imported files load at session start,
so CLAUDE.md + AGENTS.md cost exactly what one merged file would. The
split exists only so other engines (which read AGENTS.md) see the same
rules. If Claude Code is your only engine, you may rename AGENTS.md to
CLAUDE.md and delete this file - one file is simpler than two.
2. Do not list your skills or subagents here. Claude Code discovers them
from .claude/skills/ and .claude/agents/ by their own descriptions, so a
table here is an always-loaded copy that drifts.
Claude-only machinery worth using instead of growing the canon:
- .claude/rules/*.md with a `paths:` frontmatter - loaded only when a
matching file is read. Put area-specific tacit knowledge there.
- A CLAUDE.md inside a subdirectory - loaded when Claude works in it.
- Hooks in .claude/settings.json - for rules that must be enforced, not
remembered (see harness-kit examples/hooks/).
Add below only rules that are true for Claude Code and false for every
other engine. Most repos need nothing below this comment.
-->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.