Stats
Actions
Tags
From research-loop
Use when the user asks to resume, load, restore, or continue a research loop from existing .research state
How this skill is triggered — by the user, by Claude, or both
Slash command
/research-loop:research-loop-resumeThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Load `.research/` state and inject a compact research context into the conversation.
Load .research/ state and inject a compact research context into the conversation.
.research/IDEA.md exists..research/tree.md exists..research/experiments/ exists, possibly empty.If state is missing, stop and tell the user to initialize a research loop first.
.research/IDEA.md..research/tree.md and extract all hypotheses, status values, and evidence IDs..research/experiments/*.md. For each experiment, extract only:
.research/DASHBOARD.md if present.# 研究上下文
## 研究动机
[IDEA one-liner plus short background]
## 假设树
- H1: [description] (被支持, Evidence: E001)
- H1.1: [description] (进行中)
- H2: [description] (待验)
## 实验历史
- E001 ([date]): 验证 H1, 结论=[support/refute/uncertain summary]
## Recent Validations
- [date] H1 [被支持]: [summary]
## Next Steps
1. [next step from DASHBOARD or analysis]
## State Files
IDEA: [absolute path]/.research/IDEA.md
Tree: [absolute path]/.research/tree.md
Dashboard: [absolute path]/.research/DASHBOARD.md
DASHBOARD.md is absent, continue without it and say so.tree.md has invalid status vocabulary, surface the format error instead of guessing.npx claudepluginhub hamondyan/research-loopCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.