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/superdeman/cockpit-agent/claude-mdgit clone --depth 1 https://github.com/SuperdeMan/cockpit-agentWrote 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/superdeman/cockpit-agent/claude-md)<a href="https://agentmods.dev/instructions/superdeman/cockpit-agent/claude-md"><img src="https://agentmods.dev/badge/instructions/superdeman/cockpit-agent/claude-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 | $0.02820 | $0.02820 |
| Opus 5 | $0.01410 | $0.01410 |
| Sonnet 5 | $0.00564 | $0.00564 |
| Haiku 4.5 | $0.00282 | $0.00282 |
Grade A, and why
cockpit-agent CLAUDE.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 4d 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.
CLAUDE.md — 智能座舱 Multi-Agent 项目规则
本文件是项目最高工程约定。调整规范时先改本文档，再改实践。 架构唯一真相源：
docs/architecture/cockpit-agent-architecture.md。
1. 项目是什么
云边协同的智能座舱 Agent 系统：端侧快系统处理高频、确定、安全敏感指令；云侧 Planner 编排复杂、多域、多轮任务。Agent 统一使用 gRPC 契约 + Manifest，经 Registry 发现。
2. 技术栈
| 层 | 主要技术 |
|---|---|
| Gateway | Go，grpc-go，WebSocket |
| Orchestrator / Agent / AI 服务 | Python 3.12，grpcio，FastAPI |
| HMI | React + TypeScript + Vite |
| Mobile | React Native + Expo + TypeScript |
| 服务通信 | gRPC；proto/ 是唯一契约源 |
| 广播 / 存储 | NATS；Redis；PostgreSQL + pgvector |
3. 目录约定与能力流程
| 目录 | 职责 |
|---|---|
proto/ |
gRPC 真相源；先改 proto，再 codegen |
gateway/ |
edge/cloud 接入网关 |
orchestrator/edge/ |
FastIntent、端侧编排、VAL 与车控知识库 |
orchestrator/cloud/ |
Planner、Context、Loop、Executor、Aggregator |
llm-gateway/ |
所有 LLM/ASR/TTS/S2S/视觉调用出口 |
registry/ |
Agent 注册、发现与 PostgreSQL round-trip |
memory/ |
记忆、画像、关系与建议准入 |
agents/ |
各领域 Agent；agents/_sdk 是公共 SDK |
skills/ |
Planner guides、policies、exemplars |
runtime/ |
跨服务唯一判据与运行时公共件 |
security/ |
权限、scope、审核、注入防护 |
payment-gateway/ |
支付网关；Agent 不持支付凭证 |
proactive/ |
主动消息全局治理 |
observability/ |
事件、collector、trace、日志、指标 |
hmi/ / mobile/ / dashboard/ |
三个客户端/观测界面 |
deploy/ |
Compose、Helm、K8s；不是 .env 真相源 |
scripts/ |
codegen、构建、部署、迁移、探针 |
test/ |
E2E、评测基线、旅程、CDP |
docs/ |
架构、设计、review、指南、历史 |
gen/ / certs/ / models/ |
生成物或本地资产，均 gitignore，不手改/提交 |
关键声明只允许一份权威：
- 车控对象/操作/风险/端侧意图：
orchestrator/edge/knowledge/commands.yaml； - 重试规则：
orchestrator/cloud/retry_policy.py； - D0/T2 流式状态：
orchestrator/cloud/stream_state.py； - 槽值形状：
orchestrator/cloud/slot_shape.py； - 候选集问答：
orchestrator/cloud/candidate_query.py； - 时间、极性、问句、安全信号等跨服务判据：
runtime/； - Planner 软知识：
skills/； - 商户能力/工具：
agents/mcp_bridge/servers.yaml。
不要在第二个模块复制判据、词表或声明；跨进程消费用测试对账声明源。
新增能力的标准流程
新增 Agent
- 在
agents/<name>/建 manifest、源码、README、tests；参考现有同类 Agent。 - Manifest 声明 intent、权限、trust、deployment、latency、context scopes 与 capability 元数据。
- 继承
agents/_sdk::BaseAgent，不重新实现 gRPC 契约。 - 在部署配置注册服务；通过 Registry 发现，不改 orchestrator 核心路由分支。
- 补契约测试、黄金用例、门禁与全量验证。
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.
- 4d ago First seen · 188 lines · 2,820 tokens per session scan A af6197a00660
cockpit-agent CLAUDE.md is an instructions file published in the GitHub repository SuperdeMan/cockpit-agent (15 stars, last pushed today), licensed Apache-2.0. It adds 2,820 tokens to every session, about $0.0141 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.
Other instructions, from other repositories
conductor AGENTS.md
AGENTS.md instructions for conductor-oss/conductor, covering agents.md, project overview, setup commands, java version references and code style.
conductor CLAUDE.md
Claude Code instructions for conductor-oss/conductor, covering claude.md — conductor (server repo), writing documentation, for each content type, start here, when you can't verify and key source locations.
hoop CLAUDE.md
Claude Code instructions for hoophq/hoop, covering claude.md, project overview, toolchain & prerequisites, architecture and module breakdown.
yorkie CLAUDE.md
Claude Code instructions for yorkie-team/yorkie, covering yorkie — crdt-based collaboration server, commands, integration test environment, commit messages and key architecture.
OGAM AGENTS.md
AGENTS.md instructions for off-grid-ai/OGAM: All rules for this repository live in rules.md — the single, tool-neutral source of truth. This file is only a pointer so the rules stay identical no matter which agent reads them.
mockd CLAUDE.md
Claude Code instructions for getmockd/mockd, covering mockd — ai agent instructions, what is mockd?, mcp server, ports and cli quick reference.