What problem does it solve?
This Skill solves the problem of receiving “deep research” outputs that are hard to audit or reproduce by producing a transparent, evidence-backed investigation process with explicit user checkpoints.
Core Features & Use Cases
- Mandated multi-cycle research per theme: Runs at least two full research cycles for each major theme (broad landscape discovery and then a targeted deep dive).
- User-controlled stop points: Requires you to approve the scope first, then approve the research plan, and finally review the completed report.
- Evidence hierarchy and confidence labels: Uses an evidence hierarchy (e.g., meta-analyses > RCTs > observational > expert opinion) and assigns confidence annotations to conclusions.
- Academic-grade writing and citations: Produces flowing narrative prose with APA 7th-style citations integrated into the argument, including methodology, limitations, and references.
- Self-contained, reproducible workflow: Uses native OpenClaw tools (such as web_search, web_fetch, and sessions_spawn) rather than relying on opaque external API wrappers.
Quick Start
Ask for a deep, exhaustive literature-style analysis by saying: Deep research on "the effects of X on Y", ensuring APA 7th citations, explicit evidence hierarchy, and three approval checkpoints.