researchvault
par lraivisto
DevOps & CloudRésumé
Local-first research orchestration engine.
Description
Ce que fait ce skill
ResearchVault is a local-first research orchestration engine that manages persistent state, synthesis, and autonomous verification for agents. All data stays in a local SQLite database with no cloud sync. Background services like MCP server and Watchdog are opt-in and must be started manually.
Unlike cloud-based research tools, ResearchVault keeps all data local with strict SSRF hardening and no background network activity unless explicitly triggered.
Quand l'utiliser
- Tracking research progress across multiple AI trend reports
- Ingesting and synthesizing content from a set of URLs into a named project
- Running autonomous strategy sessions on collected research data
- Hosting a local web portal to browse and manage research vaults
- Verifying and cross-referencing ingested sources before synthesis
Exemple de workflow
User asks: Research and synthesize AI trends for 2026
- 1
Initialize a new vault project with an objective and name
- 2
Ingest relevant URLs into the project using the scuttle command
- 3
Run the strategy command to generate an autonomous research plan
- 4
Optionally start the portal to review results in a browser UI
- 5
Review synthesized output stored in the local SQLite database
A structured research vault with ingested sources, synthesis, and strategy recommendations stored locally
Prérequis
- BRAVE_API_KEY environment variable (optional, for Brave Search)
- SERPER_API_KEY environment variable (optional, for Serper search)
- SEARXNG_BASE_URL environment variable (optional, for SearXNG search)
Installation
clawhub install lraivisto/researchvaultExécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.