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/walvez/grok-subagent/agents-mdgit clone --depth 1 https://github.com/Walvez/grok-subagentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/walvez/grok-subagent/agents-md)<a href="https://agentmods.dev/instructions/walvez/grok-subagent/agents-md"><img src="https://agentmods.dev/badge/instructions/walvez/grok-subagent/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00591 | $0.00591 |
| Opus 5 | $0.00296 | $0.00296 |
| Sonnet 5 | $0.00118 | $0.00118 |
| Haiku 4.5 | $0.00059 | $0.00059 |
Grade A, and why
grok-subagent AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
AGENTS.md
项目定位
Codex 插件：通过官方 Grok Build CLI，把 Grok 作为受控外部子 Agent，并支持仓库外的 Grok 原生 X/Web 搜索。Codex 负责编排与最终核验。
怎么跑起来
# 依赖：Node.js 22+、已登录的官方 Grok CLI（~/.grok/bin/grok 或 grok）
npm test
codex plugin marketplace add "$PWD"
codex plugin add grok-subagent@walvez-grok
# 安装后必须新建 Codex 任务，旧会话不会热加载 Skill/MCP
搜索 smoke（会消耗 Grok 额度）：
python3 plugins/grok-subagent/scripts/run_search.py run --platform x --depth quick --since 3d "..."
技术栈
- 无 npm 运行时依赖
- MCP stdio bridge：
plugins/grok-subagent/mcp-server/server.mjs - 官方
grok agent stdio（ACP）用于只读/写入/交互 - 隔离搜索桥：
plugins/grok-subagent/scripts/run_search.py（改编自 sudoHG/codex-grok-search）
目录与约定
- 插件根：
plugins/grok-subagent/ - Skill：
plugins/grok-subagent/skills/grok-subagent/SKILL.md - 中文 README 是默认入口；英文见
README.en.md；README.zh-CN.md仅为迁移跳转 - 用户可见行为变更时同步更新
README.md、README.en.md、CHANGELOG.md、ARCHITECTURE.md、SECURITY.md - 写入模式只允许 linked Git worktree（
.git为文件），禁止主检出 - 搜索模式禁止进入当前仓库 cwd；结果缓存在
~/.cache/grok-subagent/search-runs - 不要提交
~/.grok/auth.json、token、私有 prompt 或含真实密钥的测试仓
当前状态与下一步
- 现役版本：
0.4.0，main已含隔离搜索工具 - 本地插件缓存：
~/.codex/plugins/cache/walvez-grok/grok-subagent/0.4.0 - 公开 GitHub Release 仍停在
v0.3.1；若要对外安装路径解析到 0.4.0，需要补 tag/Release - 新任务中优先：项目审查用
grok_spawn_readonly；X/Reddit/实时公开研究用grok_search
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 · 46 lines · 591 tokens per session scan A 67e356c79db6
grok-subagent AGENTS.md is an instructions file published in the GitHub repository Walvez/grok-subagent (4 stars, last pushed 1mo ago), licensed MIT. It adds 591 tokens to every session, about $0.0030 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
pi-go CLAUDE.md
Claude Code instructions for dimetron/pi-go, covering claude.md — pi-go, work in a git worktree, not the primary checkout, where worktrees live, what pi-go's agent already does, and why it matters and commits: all commits must be signed.
peri CLAUDE.md
Claude Code instructions for KonghaYao/peri, covering claude.md — perihelion, 先读什么, 跨仓库契约, 任务路由 and 代码定位.
grok-keysmith AGENTS.md
AGENTS.md instructions for Jia-Ethan/grok-keysmith, covering repository agent guide, scope, release and external state, change boundaries and documentation rules.
pi-xai-oauth AGENTS.md
AGENTS.md instructions for BlockedPath/pi-xai-oauth, covering agents.md — ai agent operations manual for pi-xai-oauth, project overview, key commands (exact, copy-paste ready), architecture & boundaries (must / must not) and file structure & wayfinding.
openinterpreter AGENTS.md
Instructions for openinterpreter/openinterpreter, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
FylloCode AGENTS.md
AGENTS.md instructions for Fioooooooo/FylloCode, covering agents.md, 项目概述, 技术栈, 目录结构 and 常用命令.