vllm.cpp AGENTS.md

A project instruction guide for contributors to vllm.cpp, a C++ implementation that follows the behavior and performance goals of vLLM. It explains required startup, role-claiming, issue, specification, and worktree procedures.

In plain words
What is it for?
Use it at the start of work on vllm.cpp to inspect the live project position, declare an appropriate role, and follow the project’s contribution rules.
Why use it?
It gives coding agents a defined process for understanding the current project state and coordinating changes safely.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/mudler/vllm.cpp/agents-md
Clone the repo
git clone --depth 1 https://github.com/mudler/vllm.cpp

Made for: Codex, OpenCode.

Per session 8,119 This file is loaded in full into every session.
When invoked 8,119 The same file — it is already loaded in full.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce 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

Measured 2d ago against content hash dfe92dd5fd6e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

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
AGENTS.md · 633 lines

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

  1. Run scripts/agent-start.py. Pass --intent operator|helper|read-only and --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.
  2. Declare a role. Use scripts/agent-role.py claim operator for a multi-step integration campaign. Use claim helper --row <ID> for one scoped task. Use claim read-only for inspection. The operator claim records the current worktree as a coordinator. Another coordinator does not block the claim. Add --headless only when the developer explicitly says the run is unattended. Never infer this setting.
  3. Run scripts/now.py to get the live position. Read .agents/NOW.md for the operator's current gate and next actions. The command output is derived. The file is authored and fits on one screen.
  4. Read only the claimed row, its spec, its evidence, and the task guide for the current job.
  5. Run scripts/agent-preflight.sh before 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.

Read the full file on GitHub · 633 lines

Changes

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.

  1. 2d ago First seen · 633 lines · 8,119 tokens per session scan A dfe92dd5fd6e

Subscribe to this mod's changes

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.