Archetype-talk
Copilot instructions2 starsChanged 47 days agoMITView on GitHub
This repo has one ruleset for every AI agent and model. It is not Claude-specific: the .claude/ folder name is historical, not a scope. If you are Cline, Roo, Continue, Copilot, Qwen, Codex, Gemini or anything else, all of it applies to you. 1. Read AGENTS.md — the cross-tool entry point. 2. Then CLAUDE.md — the full working rules: grounding and Levels, Signal vs Evidence, never inventing quotes/statistics/sources, demo-data isolation, link changes only after the user approves them. 3. Workflows…
Where files are headingNew Copilot instructions files have more rules about branches and pull requests, and fewer about code style.
Of the Copilot instructions files we read, the 2,343 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 repository31 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- affinitySkill
- ai-personaSkill
- analytics-syncSkill
- backlogSkill
- catalogSkill
- cold-startSkill
- contradictionsSkill
- demo-dataSkill
- dovetail-syncSkill
- exportSkill
- extract-findingsSkill
- feature-panelSkill
- forget-participantSkill
- graph-lintSkill
- interview-guideSkill
- journey-mapSkill
- opportunity-treeSkill
- persona-avatarSkill
- persona-querySkill
- persona-talkSkill
- persona-voiceSkill
- persona-workshopSkill
- prdSkill
- prioritizeSkill
- researcherSkill
- retentionSkill
- source-syncSkill
- subject-accessSkill
- welcomeSkill
The file
We split this copilot-instructions.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
- Distinctive: line 8
Read AGENTS.md — the cross-tool entry point.
- Very specific: line 27
Never read or edit app/index.html or app/src/embedded-graph.html — 300 KB+ generated build artifacts.
≈ “Common” is an estimate: at least 1 in 5 of the 3,779 Copilot instructions 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
Archetype Talk — read this before doing anything
This repo has one ruleset for every AI agent and model. It is not
Claude-specific: the .claude/ folder name is historical, not a scope. If you
are Cline, Roo, Continue, Copilot, Qwen, Codex, Gemini or anything else, all of
it applies to you.
- Read AGENTS.md — the cross-tool entry point.
- Then CLAUDE.md — the full working rules: grounding and Levels, Signal vs Evidence, never inventing quotes/statistics/sources, demo-data isolation, link changes only after the user approves them.
- Workflows are plain instructions in
.claude/skills/<name>/SKILL.md; the index is .claude/skills/INDEX.md. Claude Code, Copilot and Cursor list them as slash-commands (/persona-talk); Codex and Gemini CLI find the same files through.agents/skills/(Codex:$persona-talk). In any other agent, open the matchingSKILL.mdand follow its steps by hand. Same result. - Follow a skill silently — never narrate it. The user gets the result in
the format that skill defines: not your plan, not your step numbers, not a
summary of what a file held before you used it. VS Code custom modes and local
reasoning models emit
<think>/<reasoning>tags as plain text the client does not strip — inpersona-talkthat prints the persona's cheat sheet above her own answer and the interview is over. Can't hide the tags? Keep what is inside them to a few words. - Fresh checkout, first session: no other tool runs Claude Code's SessionStart
hook, so do its job yourself — check
Inbox/for unprocessed files, and if.claude/onboarded.localis missing, follow.claude/skills/welcome/SKILL.md. - Never read or edit
app/index.htmlorapp/src/embedded-graph.html— 300 KB+ generated build artifacts. Edit the partials inapp/src/(map: app/src/README.md) and rebuild withpython3 scripts/build_app.py.
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.