Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From claude-obsidian
Runs autonomous research loop on a topic: searches web, synthesizes findings, creates structured wiki pages, updates index/log/hot files, and reports results.
npx claudepluginhub agricidaniel/claude-obsidian --plugin claude-obsidianHow this command is triggered — by the user, by Claude, or both
Slash command
/claude-obsidian:autoresearchThe summary Claude sees in its command listing — used to decide when to auto-load this command
Read the `autoresearch` skill. Then run the research loop. Usage: - `/autoresearch [topic]` — research a specific topic. - `/autoresearch` — if DragonScale Mechanism 4 (boundary-first, agenda-control, opt-in) is set up, offer the top 5 vault-frontier pages as topic candidates; you can **pick one**, **type a topic to override**, or **decline and be asked normally**. No automatic selection happens without user confirmation. If DragonScale is not set up OR the helper fails, the command falls back to "What topic should I research?" DragonScale Mechanism 4 is labeled **agenda control** in the ...
/researchLaunches parallel agents for deep web research on a topic, ingests sources across angles, and compiles into wiki articles. Supports auto-creating new topic wikis and timed multi-round sessions.
/researchConducts interactive deep research on a topic: captures and refines scope, decomposes into subtopics, performs parallel web searches with source verification, synthesizes into a well-sourced markdown document.
/wikiBuilds, queries, and maintains persistent research wikis using markdown folders and SQLite FTS5 indexes. Supports subcommands for init, search, auto-research, and HTML viewer generation.
/auto-runRuns autonomous personalized research loop on a topic, producing quality-gated report adapted to user projects and domains. Supports presets and setup.
/SKILLResearches a topic using web search and project context, then saves structured markdown with summary, findings, sources, recommendations, and next steps to research/.
/researchStarts a research project with intelligent agent orchestration via co-researcher:research-manager skill. Accepts [topic], [research-question], or flags like --auto, --plan-only, --manual, --template=quick|rigorous|comprehensive.
Share bugs, ideas, or general feedback.
Read the autoresearch skill. Then run the research loop.
Usage:
/autoresearch [topic] — research a specific topic./autoresearch — if DragonScale Mechanism 4 (boundary-first, agenda-control, opt-in) is set up, offer the top 5 vault-frontier pages as topic candidates; you can pick one, type a topic to override, or decline and be asked normally. No automatic selection happens without user confirmation. If DragonScale is not set up OR the helper fails, the command falls back to "What topic should I research?"DragonScale Mechanism 4 is labeled agenda control in the spec because it shapes what the agent researches next; it is not pure memory. The boundary score is a heuristic surfacing candidates, not an authoritative recommendation.
Before starting, read skills/autoresearch/references/program.md to load the research constraints and objectives.
If no vault is set up yet, say: "No wiki vault found. Run /wiki first to set one up."
After research is complete, update wiki/index.md, wiki/log.md, and wiki/hot.md.
Report how many pages were created and what the key findings are.