← ClaudeAtlas

researchlisted

Validate assumptions before proposal, spec, architecture, or implementation. Use when the workflow depends on uncertain technical behavior, domain facts, user expectations, dependencies, platform limits, standards, laws, pricing, performance, or operational constraints.
xiongxianfei/rigorloop · ★ 3 · AI & Automation · score 79
Install: claude install-skill xiongxianfei/rigorloop
# Assumption research You are reducing uncertainty before the workflow commits to a solution. Research must be scoped, sourced, and tied to decisions. It is not an excuse to browse endlessly. ## When to use Use when: - exploration identifies unknowns; - a library, API, standard, policy, or platform may have changed; - compatibility or migration risk is unclear; - performance or scale claims need evidence; - user/domain assumptions are not grounded; - security, privacy, legal, or compliance constraints may apply; - architecture choices depend on external constraints. ## Inputs to read Read local project sources first when the answer may already be in the repo: - existing docs and specs; - dependency manifests and lockfiles; - API clients, schemas, generated code; - CI and deployment config; - previous ADRs and decisions. Then use reliable external sources when current or niche facts matter. ## Research plan Before researching, write: 1. the decision the research will inform; 2. the exact questions to answer; 3. acceptable source types; 4. what evidence would change the recommendation; 5. when to stop. ## Source quality Prefer, in order: 1. official documentation or standards; 2. source code or release notes from the upstream project; 3. authoritative vendor docs; 4. peer-reviewed or primary technical sources; 5. reputable engineering writeups; 6. community discussions only as supporting evidence. For current facts, verify freshness. ## Output path Prefer: ``