Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/mudler/vllm.cpp/agents-mdgit clone --depth 1 https://github.com/mudler/vllm.cppWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.08119 | $0.08119 |
| Opus 5 | $0.04059 | $0.04059 |
| Sonnet 5 | $0.01624 | $0.01624 |
| Haiku 4.5 | $0.00812 | $0.00812 |
Grade A, and why
vllm.cpp AGENTS.md scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
container carrying `git`, `curl`, `wget`, `gcc`, `cmake`, `ninja`, `python3` and How it starts
The opening of the file, as written. The whole thing — 633 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md: the rules
This file contains the complete policy for vllm.cpp. It is the only file that
every agent loads automatically, so every rule lives here. Files under
.agents/ are task guides. They explain how to do a specific job. They cannot
add or weaken a rule in this file.
The project mirrors vLLM in C++ without PyTorch or a ggml dependency. vLLM defines the reference behavior and the performance target.
Start here
- Run
scripts/agent-start.py. Pass--intent operator|helper|read-onlyand--row <ID>when you know them. Otherwise, relay its welcome and ask what work is intended. Follow the printed action, then run the command again. - Declare a role. Use
scripts/agent-role.py claim operatorfor a multi-step integration campaign. Useclaim helper --row <ID>for one scoped task. Useclaim read-onlyfor inspection. The operator claim records the current worktree as a coordinator. Another coordinator does not block the claim. Add--headlessonly when the developer explicitly says the run is unattended. Never infer this setting. - Run
scripts/now.pyto get the live position. Read.agents/NOW.mdfor the operator's current gate and next actions. The command output is derived. The file is authored and fits on one screen. - Read only the claimed row, its spec, its evidence, and the task guide for the current job.
- Run
scripts/agent-preflight.shbefore you edit a file.
Never infer a role, host, permission, or developer preference. Resolve .env
and .agents/developer-preferences.md from the shared checkout. Ask only for
the one value that the current gate needs. If a value is unavailable, leave its
gate PENDING. Never convert a missing value into an assumption. Preferences
control operations only. They cannot reduce a correctness, evidence,
attribution, or testing obligation.
Create both files on first use. Neither is tracked, so a fresh checkout has
neither, and scripts/agent-start.py reports the absence and routes you to ask.
Ask the developer for the one value the current gate needs. Record an
environment value with scripts/agent-onboard.py --env-set KEY=VALUE, which
refuses any key .env.example does not declare. Record a preference by copying
.agents/developer-preferences.example.md and editing the one entry. Leave
every key you did not ask about empty, because empty means unavailable and its
gate stays PENDING. A host name, a share path, or a checkout path written in
a repository document is another developer's resolved value. It is never a
default, and reading one instead of asking is the failure this rule names.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 633 lines · 8,119 tokens per session scan A dfe92dd5fd6e
vllm.cpp AGENTS.md is an instructions file published in the GitHub repository mudler/vllm.cpp (367 stars, last pushed 2d ago), licensed Apache-2.0. It adds 8,119 tokens to every session, about $0.0406 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
osaurus AGENTS.md
Instructions for osaurus-ai/osaurus, covering codex configuration - osaurus-staging, build & test, keychain tip (optional), osaurus release proof and pr reporting and model runtime non-negotiables.
semantic-router copilot-instructions.md
Instructions for vllm-project/semantic-router, covering github copilot instructions, start here and review priorities.
semantic-router AGENTS.md
Instructions for vllm-project/semantic-router, covering vllm semantic router agent entry, read first, task routing, repository map and supported environments.
openmed AGENTS.md
Instructions for maziyarpanahi/openmed, covering repository guidelines, project structure & module organization, build, test, and development commands, linting and formatting and coding style & architecture rules.
Mooncake AGENTS.md
Instructions for kvcache-ai/Mooncake, covering agents.md, docs/ directory changes and pull request guidelines.
speech-swift AGENTS.md
Instructions for soniqo/speech-swift, covering agent instructions, communication style, workflow, running tests and benchmarks — sequential and memory-aware and git conventions.