Loading...
Loading...
Found 14 Skills
Use when a claim depends on people taking a real action.
Use when finding contradictions or reasoning gaps.
Use when a claim depends on an agent taking a real action.
Read-only exploration, status checks, and reporting without modifications. Use when user asks to check status, find files, search code, show state, or explicitly requests read-only investigation. Do NOT use when user wants changes, fixes, refactoring, or any write operation.
Read-only GitHub triage for issues AND PRs. 1 item = 1 background task (category: quick). Analyzes all open items and writes evidence-backed reports to /tmp/{datetime}/. Every claim requires a GitHub permalink as proof. NEVER takes any action on GitHub - no comments, no merges, no closes, no labels. Reports only. Triggers: 'triage', 'triage issues', 'triage PRs', 'github triage'.
This skill is used when the answer requires web retrieval or independent multi-source verification, including fact-checking, technical comparison, current information, conflicting claims, important suggestions, as well as in-depth research that requires multiple rounds of deep exploration and iterative convergence.
AI Berkshire Skill: Investment Thesis Drift Detection: Distinguish Between Fact Changes and Wording Changes. Source: skills/thesis-drift.md.
Verify a specific claim by searching for evidence across web and academic sources. Use when the user asks to verify, fact-check, or confirm a statement.
You MUST use this when work is about to be called complete, done, fixed, ready, or mergeable - before a completion claim, a merge or pull request, a branch finish, or a handoff report.
Audit load-bearing paper claims against evidence in results, tables, figures, experiments, or citations. Use for claim audits, overclaiming checks, first or state-of-the-art claims, significant-result wording, prose-table number mismatches, and camera-ready or rebuttal evidence checks.
Discover, verify, and document literature based on specific judgments, and adjust search directions according to evidence. Use when the user asks for "find relevant literature", "conduct literature discovery", "supplement sources for this argument", or requests the rw-literature-discovery workflow. Runs without a private local workspace or preset research-lab; use user-provided material and bundled public-source methods.
Standalone skill for validating an explicit assumption with structured research. Use when a user holds a belief that is still an assumption — a technical claim, a library's behavior, a "this is probably how X works" — and wants it investigated methodically before it leaks into planning or code. Recommends a research methodology (non-binding) and returns an evidence-backed, graded verdict (Confirmed / Refuted / Inconclusive). Trigger on "structured-research", "validate this assumption", "is it true that", "research whether", "verify my assumption", "check this claim", "I assume X — is that right?". Standalone like bug-hunting / hotfix / brand-design — no pipeline required.