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/hoanganhduc/openclaw-bot --skill graph-verifier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 | graph_verifier |
| description | Verify small graph-theoretic claims using a local Python helper. |
| user-invocable | true |
| disable-model-invocation | true |
| metadata | {"openclaw":{"emoji":"📐","requires":{"bins":"[Truncated]"}}} |
Use this skill when the user asks to sanity-check a finite graph claim, inspect a small construction, or validate a graph encoding.
/tmp/graph_input.json with the graph data.exec: /workspace/skills/graph-verifier/run_graph_verifier.sh --input /tmp/graph_input.json
graph_data: NetworkX node-link JSONedges: list like [[1,2],[2,3]]adjacency: object mapping nodes to neighbor listsexpected: optional expected values such as {"connected": true, "bipartite": false}For heavy computations beyond NetworkX's capabilities, use the SageMath skill instead:
Use this graph-verifier for simple/fast checks: connectivity, bipartiteness, degree sequence, basic properties.