universal-agent-seed
CLAUDE.md0 starsChanged 4 months agoMITView on GitHub
Read AGENTS.md first — it is the single source of truth for commands, structure, code style, git workflow, and boundaries. This file only adds Claude-Code-specific wiring. - Knowledge is auto-loaded at session start by the SessionStart hook (.claude/hooks/inject_knowledge.py). You don't need to re-read .claude/knowledge/* manually unless you need a specific detail. - When the session banner reports pending learnings, run /reflect before major work. - Use the curated MCP servers: Context7 for any library API (never guess versions), Serena for…
- Reads credentials
In this repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/00-defer-to-agents-md.mdcCursor rule
- bootstrap-stackSkill
- distillSkill
- extract-versionSkill
- learnSkill
- reflectSkill
- seed-updateSkill
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 17
When the session banner reports pending learnings, run /reflect before major work.
- Distinctive: line 31
Keep this file small (< ~2,500 tokens).
≈ “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
CLAUDE.md
Read AGENTS.md first — it is the single source of truth for commands, structure,
code style, git workflow, and boundaries. This file only adds Claude-Code-specific wiring.
@AGENTS.md
<!-- core-rules.md is intentionally NOT @-imported here: the SessionStart hook
(inject_knowledge.py) already injects it (plus the knowledge store and the pending-learnings
nudge). Importing it too would duplicate it in context. One channel — the hook — owns it. -->Claude-specific notes
- Knowledge is auto-loaded at session start by the
SessionStarthook (.claude/hooks/inject_knowledge.py). You don't need to re-read.claude/knowledge/*manually unless you need a specific detail. - When the session banner reports pending learnings, run
/reflectbefore major work. - Use the curated MCP servers:
Context7for any library API (never guess versions),Serenafor navigating large code,Playwrightto verify UI,GitHubfor repo/PR work. - Growth entry points:
/learn(record a fact now),/reflect(promote pending learnings),/distill(turn a finished task into a skill),/bootstrap-stack(add a language). - Safety hooks block secret/
.envaccess and destructive commands. Do not work around them — if one fires, fix the approach.
Memory map
- Durable project facts →
.claude/knowledge/ - Behavior rules (precise, behavior-changing) →
.claude/rules/core-rules.md - Reusable procedures →
.claude/skills/ - Keep this file small (< ~2,500 tokens). Put detail in imported/linked files.
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.