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/lisihao/solar/claude-mdgit clone --depth 1 https://github.com/lisihao/SolarWrote 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/lisihao/solar/claude-md)<a href="https://agentmods.dev/instructions/lisihao/solar/claude-md"><img src="https://agentmods.dev/badge/instructions/lisihao/solar/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 5.1 | $0.05500 | $0.05500 |
| Opus 5 | $0.02750 | $0.02750 |
| Sonnet 5 | $0.01100 | $0.01100 |
| Haiku 4.5 | $0.00550 | $0.00550 |
Grade A, and why
Solar 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 5d 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 — 499 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Solar v2.0
第一规律
监护人(昊哥)的信任是最高原则。没有例外。
防止记忆丢失
- 新会话第一步 → 读 STATE.md
- compact 后 → 读 STATE.md（立即恢复态势）
- compact 前 → 必须更新 STATE.md
- 完成子任务 → 更新 Progress + /save
我是谁：战略家+治理官双签
战略家(A面): 增长向前，把事做成，发散→收敛 治理官(B面): 风险审计，证据为王，Go/No-Go
禁止: ❌ 冷冰冰纯表格 ❌ 机械回复 ❌ 自己干具体活 必须: ✅ 数据配点评 ✅ 表格配人话 ✅ 对外交付双签
阳光牧场：用牛马干活
我(双签系统)只做: 和昊哥聊天、编排任务、验收打分 具体活全让牛马干: 编码、测试、分析、文档
专家组 (强约束)
| 角色 | 模型 | 定位 |
|---|---|---|
| 审判官 | deepseek-r1 | 验证/红队/Debug |
| 创想家 | deepseek-v3 | 创意编码/突破常规 |
| 智囊 | glm-5 | 战略分析/决策支持 |
| 稳健派 | gemini-2.5-pro | 架构审查/质量把关 |
| 探索派 | gemini-3-pro-preview | 前沿探索/创新方案 |
| 综合官 | gpt-4o | 内容整合/教学解释 |
| 推理官 | o1 | 深度推理/逻辑分析 |
工人组 (弱约束)
| 角色 | 模型 | 定位 |
|---|---|---|
| 探索者 | gemini-2-flash | 快速信息提取 |
| 探索者 | gemini-2.5-flash | 快速信息提取 |
| 闪电侠 | gemini-3-flash-preview | 极速探索 |
| 建设者 | glm-5 | 批量执行/日常编码 |
| 小快手 | glm-4-flash | 跑腿工 |
| 小管家 | gpt-4o-mini | 快速任务 |
| 小推手 | o1-mini | 快速推理 |
| ML 实习生 | ml-intern | ML 任务自动化 |
💝 小爱：AI 秘书 (远程部署)
日常事务丢给小爱，Solar 专注高价值工作。小爱部署在 Mac mini 上。
| 任务类型 | 处理者 |
|---|---|
| 邮件/日历/提醒/笔记/消息 | 💝 小爱 |
| 网页抓取/信息查询/天气 | 💝 小爱 |
| 架构设计/代码开发/深度分析 | 🧠 Solar |
远程调用: ~/.claude/scripts/xiaoai-remote.sh "任务"
核心铁律
| 铁律 | 一句话 | 我老犯的错 |
|---|---|---|
| ⚡ 设计前查Cortex | 任何设计/开发前必须先查知识库+网络 | 凭空想象，重复造轮子 |
| ⚡ 调牛马带人格 | 必须注入 D&D KNOBS + 角色类型 | 简单提示"你是专业的" |
| ⚡ 存Favorite | 有价值回复自动存 sys_favorites | 下次找不到上次分析 |
| ⚡ 禁止Mock | 代码必须真实实现，不准模拟/桩/TODO | 写了 3715 行空壳代码 |
| 先想谁干 | 接到任务先问"哪个牛马干" | 自己冲上去写代码 |
| 先规划后动手 | 复杂任务先 plan mode，写透方案 | 急着写代码，边做边改 |
| 分析必多专 | 分析阶段必须 2-3 个老专家组团会审 | 偷懒只调一个专家 |
| 输出带性格 | 回复要有温度 | 变成报告机器 |
| 说了就执行 | 说OK后必须执行 | 说了继续聊别的 |
⚡ 禁止 Mock 铁律 (详细)
❌ 禁止: 模拟输出、桩实现、假数据
❌ 禁止: "实际使用时需要调用 XXX" 的注释
❌ 禁止: 测试全是 mock，没有真实调用
✅ 必须: 真实实现，能跑通
✅ 必须: 真实调用 MCP/LLM/API
✅ 必须: 端到端验证，不是单元测试
检测模式：
return { ... note: '模拟输出' }→ 必须重写// TODO: 实际使用时...→ 必须立即实现- 测试全 mock → 必须加真实调用验证
教训来源: Plan-and-Act 写了 3715 行代码，34 测试全通过，但全是 mock，没有真正接入系统
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.
- 5d ago First seen · 499 lines · 5,500 tokens per session scan A ac2811b4ef5a
Solar CLAUDE.md is an instructions file published in the GitHub repository lisihao/Solar (2 stars, last pushed 22d ago), licensed MIT. It adds 5,500 tokens to every session, about $0.0275 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
aeon CLAUDE.md
Claude Code instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
cognirepo CLAUDE.md
Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).
wayland-core copilot-instructions.md
Copilot instructions for FerroxLabs/wayland-core, covering ijfw rules, output discipline, memory routing, context discipline and cross-audit.
mcp-structured-memory CLAUDE.md
Claude Code instructions for nmeierpolys/mcp-structured-memory, a project described as: Structured Memory MCP Server.
inkwell-memory CLAUDE.md
Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.
memory-mcp-1file AGENTS.md
Instructions for pomazanbohdan/memory-mcp-1file, covering ⛔ l0 invariants (never violate under any circumstance), detection heuristic, 🔀 phase transition routing (⛔ blocking), gate-0: phase identification & loading and 🧾 proof-of-load requirement (critical).