From qiushi-skill
Investigates a given topic by first gathering facts, distinguishing confirmed facts from reasonable inferences and open questions, then forming a judgment or recommendation supported by evidence.
How this command is triggered — by the user, by Claude, or both
Slash command
/qiushi-skill:investigation-firstThe summary Claude sees in its command listing — used to decide when to auto-load this command
读取并遵循 `skills/investigation-first/SKILL.md`。 调查任务: $ARGUMENTS 输出要求: 1. 写出调查目的和范围。 2. 区分确认事实、合理推断和存疑事项。 3. 结论必须由事实支撑。 4. 信息明显不足时,不要提前下结论。
读取并遵循 skills/investigation-first/SKILL.md。
调查任务: $ARGUMENTS
输出要求:
npx claudepluginhub hughyau/qiushi-skill --plugin qiushi-skill7plugins reuse this command
First indexed Mar 28, 2026
Showing the 6 earliest of 7 plugins
/exploreForces exploratory thinking that delays commitment and retains competing interpretations for a given question or topic.
/q-challengeRigorously challenges all investigation assumptions and hypotheses by evaluating evidence, checking for cognitive biases, and producing a structured challenge document.
/divePerforms deep-dive investigation on the specified topic using the investigation skill. Requires an active session; errors and instructs on starting one if absent.
/vault-autoresearchRuns a 3-round autonomous research loop on a topic: survey, deep dive, and cross-examination with contradictions, saving intermediate notes and a final synthesis to the vault.
/domain-researchGuides you through a 5-step conversational research pipeline for any domain, building context, generating questions, identifying gaps, extracting insights, synthesizing findings, and creating action plans.
/researchConducts systematic research via four-phase methodology (discovery, investigation, synthesis, reporting) with multi-hop reasoning, web tools, and evidence synthesis. Produces cited reports with confidence levels.