codex-acp AGENTS.md

Repository instructions for a Rust program that embeds a Codex agent and exposes it through the Agent Client Protocol, a standard way for applications to communicate with coding agents. They map the source files and describe the agent, sessions, prompts, logging, and filesystem bridge.

In plain words
What is it for?
Use them when working on the library or command-line entry point, agent sessions, slash commands, streaming prompts, logging, configuration, or the standalone filesystem service.
Why use it?
They make the Rust architecture easier to navigate and help changes fit the existing session, authentication, command, and tool-handling paths.

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/cola-io/codex-acp/agents-md
Clone the repo
git clone --depth 1 https://github.com/cola-io/codex-acp

Made for: Codex, OpenCode.

Per session 3,068 This file is loaded in full into every session.
When invoked 3,068 The same file — it is already loaded in full.
Security scan A 0 findings. 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 $0.03068 $0.03068
Opus 5 $0.01534 $0.01534
Sonnet 5 $0.00614 $0.00614
Haiku 4.5 $0.00307 $0.00307

Measured 2d ago against content hash f7cddd0a2d3b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codex-acp AGENTS.md scanned grade A with 0 findings 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 332 lines

How it starts

The opening of the file, as written. The whole thing — 332 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Repository Guidelines

This document describes how to work in this repo using idiomatic Rust patterns and the current module layout.

Project Structure

  • src/
    • lib.rs — library crate root; exposes agent, fs, logging; re-exports CodexAgent, SessionManager, FsBridge, and init_from_env for embedders.
    • main.rs — binary entrypoint; initializes tracing, loads config + profiles, boots the filesystem bridge, and wires the ACP runtime. Pass --acp-fs-mcp to run the standalone filesystem MCP server.
    • logging.rs — tracing init helpers driven by env variables (init_from_env), including optional file logging and daily rotation.
    • agent/
      • mod.rs — Agent trait façade; wires ACP trait methods into CodexAgent and re-exports ClientOp.
      • core.rs — defines CodexAgent and ClientOp, handles initialize/auth/new/load session, and applies session mode/model mutations while coordinating the auth/conversation managers.
      • prompt.rs — streaming prompt pipeline, slash-command detection, tool + reasoning updates, cancel handling, and extension method stubs.
      • commands.rs — slash command registry (AVAILABLE_COMMANDS) plus helpers like handle_slash_command and /status rendering.
      • events.rs — Codex Event → ACP update formatting (EventHandler), approval option helpers, and ReasoningAggregator for thought text.
      • config_builder.rs — builds per-session Config instances, injects filesystem guidance, and produces MCP server configs (prepare_fs_mcp_server_config, build_mcp_server).
      • session_manager.rs — session state store (SessionState, SessionManager), conversation caching, capability tracking, notification helpers, and apply_context_override.
      • utils.rs — shared helpers for session modes/models, provider detection, tool formatting (format_command_call, describe_mcp_tool, is_custom_provider, etc.).
    • fs/
      • mod.rs, bridge.rs, mcp_server.rs — filesystem bridge runtime (FsBridge::start) and MCP server entry (run_mcp_server) communicating via ACP_FS_BRIDGE_ADDR/ACP_FS_SESSION_ID.
  • Cargo.toml, rust-toolchain.toml
  • README.md, AGENTS.md
  • Makefile, scripts/stdio-smoke.sh

Read the full file on GitHub · 332 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 · 332 lines · 3,068 tokens per session scan A f7cddd0a2d3b

Subscribe to this mod's changes

codex-acp AGENTS.md is an instructions file published in the GitHub repository cola-io/codex-acp (142 stars, last pushed 7mo ago), licensed Apache-2.0. It adds 3,068 tokens to every session, about $0.0153 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.