adaptive-skill-router
AGENTS.md1 starsChanged 33 days agoMITView on GitHub
This project currently supports Codex only. Preserve the distinction between candidate retrieval and the host's semantic Skill choice. Do not turn metadata descriptions or route suggestions into automatic execution or authorization. The core purpose is choosing and coordinating already-installed skills. External search and source collection are retained optional utilities for explicitly requested discovery work, not default fallbacks when local selection finds no match. Keep entries, metadata, documentation and generated integration aligned with this scope. Explore relevant code and existing checks…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 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 repository2 other files it gives its agents.
- adaptive-skill-routerSkill
- adaptive-workflowSkill
The file
We split this AGENTS.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 11
Verify behavior changes with relevant tests and routing policy changes with npm run eval.
- Distinctive: line 13
Keep publication and performance claims factual.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.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
Development instructions
This project currently supports Codex only. Preserve the distinction between candidate retrieval and the host's semantic Skill choice. Do not turn metadata descriptions or route suggestions into automatic execution or authorization.
The core purpose is choosing and coordinating already-installed skills. External search and source collection are retained optional utilities for explicitly requested discovery work, not default fallbacks when local selection finds no match. Keep entries, metadata, documentation and generated integration aligned with this scope.
Explore relevant code and existing checks first. Clear local changes can proceed directly; use a specialized method only for an actual knowledge or method gap. The reusable guidance is in skills/adaptive-skill-router/SKILL.md when installed-skill selection or coordination remains uncertain. Preserve user goals and project constraints over broad third-party workflow triggers.
Current priority is engineering quality of shipped capabilities. Defer new features and workflow expansion; fix reproducible defects, usability failures and compatibility gaps first. Do not start further comparisons against default Codex unless the user explicitly requests them again. Preserve existing reports as historical evidence.
Verify behavior changes with relevant tests and routing policy changes with npm run eval; run npm run check for changes spanning components. Documentation edits need relevant example/link review rather than mirrored tests. Do not include local evidence, private configurations, generated integration paths or behavioral sandboxes in distributed files.
Keep publication and performance claims factual. Passing policy tests does not establish better agent quality or reduced token usage. Native discovery represents a fresh CLI process, not the exact running desktop 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.