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/alanshurafa/co-evolution/agents-mdgit clone --depth 1 https://github.com/alanshurafa/co-evolutionWhat 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 | $0.03010 | $0.03010 |
| Opus 5 | $0.01505 | $0.01505 |
| Sonnet 5 | $0.00602 | $0.00602 |
| Haiku 4.5 | $0.00301 | $0.00301 |
Grade A, and why
co-evolution 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.
How it starts
The opening of the file, as written. The whole thing — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
For agents reading this repo
This repo implements the Bounce Protocol — a convention for structured iterative refinement between AI agents using in-document markers. If you are an agent invoked against documents in this repo, you should know how to read and emit these markers.
The two markers
[CONTESTED]— appears on its own line below text you disagree with. The note states the disagreement and proposes a concrete alternative. If you see one in a document you are editing, either resolve it (delete it after applying its alternative or rebutting it) or leave it for the next agent.[CLARIFY]— appears on its own line below ambiguous text. The note offers two interpretations or asks a question with a finite answer space. If you can disambiguate, do so and delete the marker.
Both markers auto-expire after 2 passes: on the final pass of any bounce, you must resolve every remaining marker and may not introduce new ones. This rule guarantees convergence.
Full spec: BOUNCE-PROTOCOL.md. Reference implementation: agent-bouncer/.
How to interact with this repo
- If you are invoked via
/co-evolutionorbash co-evolve-bouncer.sh, you are in the general-purpose co-evolution workflow for questions, drafts, plans, specs, arguments, and markdown refinement - If you are invoked via
/dev-review(Claude Code) orbash dev-review/codex/dev-review.sh(Codex), you are inside the bounce pipeline — your role (reviewer / composer) and pass number are passed to you in the prompt template - If you are invoked via
/codex-build, you are orchestrating a detached Codex build: the session plans and reviews, Codex executes in the background under--preset codex-build, and the session is woken at gates (seeskills/codex-build/) - If you are orchestrating via
--preset claude-build, Codex plans and reviews while Claude executes the build synchronously; this path hard-requiresclaudeCLI auth because Claude is the executor (seedev-review/codex/claude-build.md) - If you are invoked via
bash agent-bouncer/agent-bouncer.sh <doc>, you are bouncing a single markdown document — same protocol applies - If you are exploring the repo directly without an explicit role, treat the protocol section above as orientation, then read the GSD-managed sections below for project meta
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 · 188 lines · 3,010 tokens per session scan A 54367aaf97f7
co-evolution AGENTS.md is an instructions file published in the GitHub repository alanshurafa/co-evolution (3 stars, last pushed 2d ago), licensed MIT. It adds 3,010 tokens to every session, about $0.0151 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 2026-08-31.
Other instructions, from other repositories
sumulige-claude AGENTS.md
Instructions for sumulige/sumulige-claude, covering agents.md - agent 系统详解, technical co-founder working agreement (always-on), layer 1: 什么是 agent？ (5 分钟), 最简单的解释 and 类比：餐厅团队.
ORCH CLAUDE.md
Instructions for oxgeneral/ORCH, covering claude.md, positioning, commands, architecture and container (di).
crewplane AGENTS.md
Instructions for crewplaneai/crewplane, covering agents.md, repo facts, read first, repo map and project heuristics.
codex-harness-javascript AGENTS.md
Instructions for thuandiep1707/codex-harness-javascript, covering codex multi-agent delivery system, system truths, public workflow registry vs internal capabilities, public workflows and internal capabilities.
crewplane CLAUDE.md
Instructions for crewplaneai/crewplane, a project described as: Run reviewable, resumable coding-agent workflows across Claude Code, Codex, Gemini, GitHub Copilot CLI, or any CLI. You define the stages in Markdown; Crewplane runs the workflow, agents do the work, and every handoff stays on disk.
crewplane GEMINI.md
Instructions for crewplaneai/crewplane, a project described as: Run reviewable, resumable coding-agent workflows across Claude Code, Codex, Gemini, GitHub Copilot CLI, or any CLI. You define the stages in Markdown; Crewplane runs the workflow, agents do the work, and every handoff stays on disk.