Evaluate whether a GitHub/open-source repository is useful for the user's selected Research KB context.
Memory
Eval Repo
Try itFetch GitHub repository metadata and project files, generate an evidence-based repository evaluation, and save the evaluation into the target knowledge base.
What it does
Fetch GitHub repository metadata and project files, generate an evidence-based repository evaluation, and save the evaluation into the target knowledge base.
The skill document
Skill: eval_repo — GitHub 仓库评估
用途
当用户发 GitHub 仓库并要求“评估/值不值得/能不能复现/环境怎么配”时,抓取 README、依赖和元信息,结合目标知识库上下文生成评估报告并保存。
触发条件
Activate when:
- 消息包含 GitHub repo 链接
- 且包含“评估、值不值、有没有用、能不能复现、环境怎么配、要不要深入”等意图
Do NOT activate when:
- 用户只是说“存一下这个 repo” →
ingest_document - 用户发 Gitea 资料仓库要求批量编译 →
batch_compile
执行流程
1. 抓取 GitHub 信息
python3 scripts/fetch_github.py --url ""
脚本返回 text_path,里面包含 README、依赖文件和仓库元信息。
2. 确定保存目标
先解析目标:
python3 scripts/resolve_eval_target.py \
--sender_id \
--chat_type \
--chat_id \
--message_sid
目标规则:
- 私聊默认个人库
- 用户明确贡献团队库时保存团队库
- 用户明确双写时保存两个库
- 群聊时必须通过
GroupSubject -> chat_bindings.json找到团队,默认保存到该团队知识库 - 群聊不允许保存到个人知识库
3. OpenClaw + MiniMax 生成评估 Markdown
必须包含:
- 三档结论:值得深入 / 选择性参考 / 暂不建议
- 与研究方向相关性
- README/依赖/活跃度/许可证/复现风险
- 与知识库已有资料关系
- 来源链接
4. 保存评估
python3 scripts/save_eval.py \
--owner --repo \
--title "<标题>" \
--eval_file /tmp/paperkb/draft_eval.md \
--brief "<简介>" \
--keywords "关键词1,关键词2" \
--score 8 \
--source_url ""
回复
回复必须给:
- 结论
- 相关性
- 复现风险
- 保存链接
脚本清单
resolve_eval_target.py:解析私聊/群聊保存目标chat_context.py:解析 OpenClaw 群聊上下文和群绑定fetch_github.py:抓取 GitHub 仓库信息save_eval.py:保存 GitHub 评估
Related skills
GitHub & developer data for AI agents — repo metadata (stars, language, license), repo audits & risk scoring, user profiles, recent releases, and gist conten...
Jointly analyze a research paper and its open-source implementation. Use when a user wants to understand a paper through code, map theory/formulas/algorithms...
Reconstruct open-source growth channels with evidence
GitHub 产品调研与机会发现 · Product research
Generate evidence-based project weekly reports from a local code repository. Use when the user asks for a project weekly report, weekly update, status summar...