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 skills/alanshurafa/co-evolution/dev-reviewnpx skills add alanshurafa/co-evolution --skill dev-reviewgit 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 | $0.00176 | $0.07773 |
| Opus 5 | $0.00088 | $0.03887 |
| Sonnet 5 | $0.00035 | $0.01555 |
| Haiku 4.5 | $0.00018 | $0.00777 |
Grade B, and why
dev-review scanned grade B 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 3d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
own `~/.codex/config.toml` defaults. How it starts
The opening of the file, as written. The whole thing — 796 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/dev-review - Plan-Bounce-Execute Workflow
Automates the workflow of composing a code plan with one AI, bouncing it between agents for refinement using [CONTESTED]/[CLARIFY] markers, then having one agent execute the final plan into code. Replaces manual copy-paste between Claude Code and Codex.
Use /co-evolution for general questions, ideas, arguments, prompt refinement,
or document bounces. Use /dev-review only when the intended deliverable is a
repo change or code-focused verification workflow.
For the detached flavor — the session plans and reviews while Codex executes
in the background and the session is woken at gates instead of babysitting — use
/codex-build (skills/codex-build/). It kicks the standalone runner with
--preset codex-build (Opus plans at high, Codex executes at xhigh, Opus
reviews at max; --verify on, bounces 2, revise-loop 1) and applies a
schema-bound ACCEPT / REVISE / ESCALATE gate. /dev-review is the interactive
single-session loop; /codex-build is the same pipeline run detached.
When --live is enabled on Windows, each Codex pass runs in a visible PowerShell
window so the user can watch progress in real time. The handoff contract stays
file-based: prompts are written to disk, Codex writes the final message to disk,
and Claude Code reads the resulting artifact before continuing.
The workflow has three phases:
- Compose - One agent creates the initial plan/approach
- Bounce - The plan document bounces between agents, each editing it directly
- Execute - The designated agent writes the final code from the refined plan
Step 1: Parse Arguments
Parse $ARGUMENTS for:
/dev-review [--composer opus|codex] [--executor opus|codex] [--bounces N|auto] [--verify] [--worktree] [--model MODEL] [--skip-plan] [--plan-only] [--live] <task description>
Defaults:
--composer opus(Opus creates the initial plan)--executor codex(Codex writes the final code)--bounces auto(bounce until convergence — zero [CONTESTED] and zero [CLARIFY] markers — up to a max of 6 passes. If not converged after 6, stop and ask the user to arbitrate the remaining markers. Use--bounces Nto override with a fixed count.)--verifyif set, adds a final code review pass after execution--modeloverrides Codex model--worktreecreates isolated worktree--skip-planjumps straight to code execution (backward compat with old behavior)--plan-onlystops after the bounce phase, outputs the refined plan--liveopens visible Windows terminal windows for Codex passes; if unavailable, warn once and fall back to the current headless behavior- Remaining text = task description
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 3d ago First seen · 796 lines · 176 tokens per session scan B 684228be8f19
dev-review is a skill published in the GitHub repository alanshurafa/co-evolution (3 stars, last pushed 3d ago), licensed MIT. It adds 176 tokens to every session and 7,773 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
recursive-task-optimizer
Build, configure, run, inspect, or troubleshoot agent-agnostic recursive improvement loops for a repository or artifact. Use when a task should be attempted repeatedly by Claude Code, Codex CLI, Hermes, or another CLI agent; when candidates must inherit mutable task instructions, a self-improving meta-procedure, and…
react-node-practices
// ❌ 错误：硬编码密钥 const apiKey = "sk-proj-xxxxx" const dbUrl = "postgres://user:pass@localhost/db".
design-audit
这是一个完整的 UI/UX 审计工作流，基于 Steve Jobs 和 Jony Ive 的设计哲学。用于对现有应用进行系统性审计，输出分阶段改进计划。.
technical-cofounder
Technical Co-Founder workflow for building a real, launchable product with strict phase gates (Discovery → Planning → Building → Polish → Handoff). Use when the user wants to build/ship an app/product/MVP for real usage or public launch.
manus-kickoff
版本: 1.0.0 作者: sumulige 标签: [workflow, planning, manuscript] 难度: 中级.
test-master
Skill "test-master" from sumulige/sumulige-claude, covering test master, 核心职责, 工作模式, 模式 1：tdd 开发 and 模式 2：e2e 测试.