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/majiayu000/spellbook/claude-code-reversenpx skills add majiayu000/spellbook --skill claude-code-reversegit clone --depth 1 https://github.com/majiayu000/spellbookWrote 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/skills/majiayu000/spellbook/claude-code-reverse)<a href="https://agentmods.dev/skills/majiayu000/spellbook/claude-code-reverse"><img src="https://agentmods.dev/badge/skills/majiayu000/spellbook/claude-code-reverse.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 | $0.00123 | $0.01645 |
| Opus 5 | $0.00062 | $0.00822 |
| Sonnet 5 | $0.00025 | $0.00329 |
| Haiku 4.5 | $0.00012 | $0.00164 |
Grade A, and why
claude-code-reverse 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 — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Local Binary Reverse
对明确指定的本机目标做证据优先的静态分析。默认只读，不执行目标，也不把“可见字符串”夸大为“已证明运行时行为”。
自主边界
- 可直接执行：读取用户指定的单个本机文件；运行
file、哈希、strings、符号表、依赖表、段表和只读反汇编工具；把分析缓存写入当前用户的私有 cache。 - 先征得同意：运行目标、附加调试器、注入/hook、解包到用户目录、安装分析工具、上传样本、分析未获授权的软件或受保护数据。
- 直接拒绝：绕过许可证、计费或访问控制，提取凭据，制作破解补丁，规避检测，或再分发专有代码/资源。
- 不确定所有权、目标路径或交付问题时先澄清，不要扫描整台机器寻找“可能的目标”。
1. 锁定目标与问题
先记录：
- 精确目标路径或命令名。
- 用户要验证的具体问题，例如 UI 文案来源、功能开关、网络端点、依赖或版本差异。
- 对比版本的两个精确文件（如适用）。
不要把进程名、应用名或猜测路径当成已确认目标。不要运行目标来“看看会发生什么”。
2. 建立文件身份
使用同目录的 extract.sh。它只接受普通文件或可由 command -v 解析的命令，缓存按内容 SHA-256 隔离并存放在 ${XDG_CACHE_HOME:-$HOME/.cache}/claude-code-reverse/。
SKILL_DIR=/path/to/claude-code-reverse
bash "$SKILL_DIR/extract.sh" info --target /absolute/path/to/target
bash "$SKILL_DIR/extract.sh" dump --target /absolute/path/to/target
bash "$SKILL_DIR/extract.sh" search --target /absolute/path/to/target "literal anchor"
bash "$SKILL_DIR/extract.sh" diff --target-a /path/to/v1 --target-b /path/to/v2 "literal anchor"
info 必须先给出规范化路径、文件类型、字节数和 SHA-256。若目标在分析期间发生变化，重新执行 info 和 dump；不要复用旧缓存作结论。
3. 按格式选择只读工具
先用 command -v <tool> 验证工具存在。缺失时报告缺失项，不要悄悄换成会执行目标的方案。
| 格式 | 首选证据 | 可选深入工具 |
|---|---|---|
| Mach-O | file、otool -L、otool -l、nm |
otool -tvV、已安装的反编译器 |
| ELF | file、readelf -h -l -d -s、objdump -x |
objdump -d、已安装的反编译器 |
| PE/COFF | file、objdump -x、ASCII/UTF 字符串 |
objdump -d、已安装的 PE 工具 |
| Wasm | file、wasm-objdump -x |
wasm2wat |
| ZIP/JAR/APK/打包资源 | 先列目录，不落盘解包 | 在用户批准的临时目录中解包后逐件分析 |
不要对不可信目标使用 ldd，因为某些实现可能通过加载器执行代码。不要对大文件直接输出完整反汇编、完整字符串或完整符号表；先用字面锚点缩小范围，把原始大输出留在 cache 或单独 artifact 中。
常见失败信号
strings几乎没有输出：目标可能被剥离、压缩、加密或只是启动器；先检查文件类型和段表，不要据此断言“没有实现”。- 刚 dump 后却提示 cache 不存在：内容哈希已经变化，说明目标文件被更新；重新执行
info和dump，不要复用旧版本结论。 - 同一字符串重复出现：universal/fat binary 可能包含多个架构切片；先按架构检查，不能把重复次数直接解释为调用次数。
- ASCII 搜索无命中：PE 或资源文件可能使用 UTF-16；确认本机
strings实现支持相应编码选项后再查，不要假定跨平台参数一致。
4. 从证据到结论
按以下顺序收敛：
- 用产品文案、端点、配置键、错误消息或符号名做字面锚点。
- 截取锚点附近上下文，并记录来源文件和 SHA-256。
- 交叉检查符号、导入依赖、相邻常量或两个版本的差异。
- 区分结论等级：
已观察（文件中直接存在）、强推断（多项静态证据一致）、未知（需要运行时或服务端证据）。
What ships with it
3 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.
- 5d ago First seen · 96 lines · 123 tokens per session scan A 46d5dea3447d
claude-code-reverse is a skill published in the GitHub repository majiayu000/spellbook (265 stars, last pushed 4d ago), licensed MIT. It adds 123 tokens to every session and 1,645 once invoked, about $0.0006 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 skills, from other repositories
tidewave-integration
Tidewave MCP runtime tools — debugging, smoke testing, live state inspection, SQL queries, hex docs. Use when evaluating code in a running Phoenix app.
elixir-idioms
OTP/BEAM patterns and Elixir idioms — GenServer, Supervisor, Task, Registry, pattern matching, with chains, pipes. Use when designing processes or debugging BEAM issues.
security
Enforce Elixir/Phoenix security — auth, OAuth, sessions, CSRF, XSS, SQL injection, input validation, secrets. Use when editing auth files, login flows, RBAC, or API keys.
michel-cli-demo-recorder
Produce proof-of-execution demos of the Packmind CLI (packmind-cli) as terminal-styled images (colors and formatting preserved exactly), for embedding in a GitHub PR. Renders a crisp master SVG and rasterizes it to a PNG — the PNG is what you embed, because GitHub does not render SVG in PR/issue bodies. Use this…
general-video
The fallback workflow for authoring custom HyperFrames video compositions at any length or format — longer or multi-scene pieces, brand / sizzle reels, montages, title cards, static loops, and freeform compositions. Input- and length-agnostic. If a specialized workflow clearly fits the input — a marketed product, a…
session-scan
Compute metrics for Claude Code sessions. Discovers via ccrider, filters trivial, computes friction/opportunity/fingerprint scores. Use for broad session triage.