vLLM-metax
AGENTS.md188 starsChanged 10 days agoApache-2.0View on GitHub
Project workflows are stored in .agents/ and can be read by any coding agent. Use this file as the discovery entry point when the client does not automatically discover .agents/skills/. Read the selected SKILL.md and follow its linked references; load only workflows relevant to the user's request. The portable coordinator role is vllm-metax-upgrade. Read it when acting as the upgrade coordinator. Clients with native subagent support may register this role in their own configuration; the Markdown file supplies its instructions.…
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 31,439 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 repository10 other files it gives its agents.
- attention-upgradeSkill
- common-senseSkill
- csrc-upgradeSkill
- environment-checkSkill
- model-adaptationSkill
- model-upgradeSkill
- patch-upgradeSkill
- registry-upgradeSkill
- vllm-metax-model-trimSkill
- vllm-metax-upgradeSkill
The file
We split this AGENTS.md into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
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 4
Use this file as the discovery entry point when the client does not automatically discover .agents/skills/.
≈ “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
Project skills and agents
Project workflows are stored in .agents/ and can be read by any coding agent.
Use this file as the discovery entry point when the client does not automatically
discover .agents/skills/. Read the selected SKILL.md and follow its linked
references; load only workflows relevant to the user's request.
| Task | Skill |
| Full upgrade or coordinated component upgrade | vllm-metax-upgrade |
| Upgrade environment preparation | environment-check |
| Shared upgrade rules and tooling | common-sense |
| New or standalone model adaptation | model-adaptation |
| Model compatibility | model-upgrade |
| Attention backends and cache contracts | attention-upgrade |
| Native operators, csrc and build contracts | csrc-upgrade |
| Registry, quantization and CustomOps | registry-upgrade |
Monkey patches in vllm_metax/patch/ | patch-upgrade |
| Reduced dummy or real-weight checkpoints | vllm-metax-model-trim |
The portable coordinator role is vllm-metax-upgrade. Read it when acting as the upgrade coordinator. Clients with native subagent support may register this role in their own configuration; the Markdown file supplies its instructions. Native agent registration formats differ between clients.
Skills follow the SKILL.md, references/, scripts/, and optional assets/
layout. Existing agents/openai.yaml files are optional client metadata; the
workflow is defined by SKILL.md. Put reusable workflow helpers and their tests
inside the owning skill; keep general project tools and generated model examples
under tools/. Shared upgrade helpers belong to
.agents/skills/common-sense/scripts/metax_upgrade/, with their CLI
guide in tooling.md.
Resolve command paths from the repository root; Markdown links are relative to their containing file. Upgrade skills require their documented environment gate for compatibility work. Maintaining these workflows or their tooling does not itself start an upgrade or require that gate.
Running linters
# Run all pre-commit hooks on staged files:
pre-commit run
# Run on all files:
pre-commit run --all-files
# Run a specific hook:
pre-commit run ruff-check --all-filesFinal upgrade acceptance requires pre-commit run --all-files to pass on the final
code. Fix all current hook failures, including pre-existing formatting issues, and
rerun until it passes. If hooks modify source, refresh affected validation evidence
before declaring the upgrade complete.
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.