princeofscale
princeofscale/bloxforge
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/princeofscale/bloxforge --skill adapters命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | adapters |
| description | Skill for the Adapters area of bloxforge. 10 symbols across 1 files. |
10 symbols | 1 files | Cohesion: 88%
evals/| File | Symbols |
|---|---|
evals/adapters/claude-mcp-adapter.ts | startServer, waitForStudio, stopServer, checkStudioConnection, listTools (+5) |
Start here when exploring this area:
startServer (Method) — evals/adapters/claude-mcp-adapter.ts:59waitForStudio (Method) — evals/adapters/claude-mcp-adapter.ts:87stopServer (Method) — evals/adapters/claude-mcp-adapter.ts:120checkStudioConnection (Method) — evals/adapters/claude-mcp-adapter.ts:133listTools (Method) — evals/adapters/claude-mcp-adapter.ts:146| Symbol | Type | File | Line |
|---|---|---|---|
startServer | Method | evals/adapters/claude-mcp-adapter.ts | 59 |
waitForStudio | Method | evals/adapters/claude-mcp-adapter.ts | 87 |
stopServer | Method | evals/adapters/claude-mcp-adapter.ts | 120 |
checkStudioConnection | Method | evals/adapters/claude-mcp-adapter.ts | 133 |
listTools | Method | evals/adapters/claude-mcp-adapter.ts | 146 |
runTask | Method | evals/adapters/claude-mcp-adapter.ts | 151 |
sleep | Function | evals/adapters/claude-mcp-adapter.ts | 252 |
approxTokens | Function | evals/adapters/claude-mcp-adapter.ts | 256 |
extractText | Function | evals/adapters/claude-mcp-adapter.ts | 261 |
factsPresent | Function | evals/adapters/claude-mcp-adapter.ts | 269 |
| Flow | Type | Steps |
|---|---|---|
Main → ExtractText | cross_community | 6 |
Main → StopServer | cross_community | 5 |
Main → Sleep | cross_community | 5 |
Main → ListTools | cross_community | 4 |
Main → ApproxTokens | cross_community | 4 |
context({name: "startServer"}) — see callers and calleesquery({search_query: "adapters"}) — find related execution flowsexplain({target: "<file or symbol>"}) — persisted taint findings (source→sink data flows), when indexed with --pdg