Manage research claim lifecycles with evidence and confidence tracking.
npx skills add https://github.com/swarm-ai-safety/swarm --skill claim
Please help me install this Agent Skill. Skill: claim Source: https://github.com/swarm-ai-safety/swarm/tree/main/.skills/claim Command: npx skills add https://github.com/swarm-ai-safety/swarm --skill claim
This Skill streamlines the process of creating, updating, and managing research claims within the SWARM Research OS, ensuring claims are well-documented, evidence-backed, and their status is clear.
Use the claim skill to create a new claim about "transaction tax above 5% reduces welfare".
High-intent search queries and answers about installing and using this skill.
Document research claims by creating claim cards that link simulation runs as evidence, assign confidence levels, and note boundary conditions for generalizability. The claim management process integrates with a file-based vault to track supporting and weakening evidence.
Manage evidence collection by adding supporting or weakening evidence to claim cards, then update confidence levels and claim status accordingly. The lifecycle management system facilitates tracking scientific propositions through a structured knowledge vault with strict evidence provenance rules.
Yes, you can update existing claims by adding new supporting or weakening evidence from simulation runs, changing the claim status, or adjusting the assigned confidence level to reflect the latest research findings within the knowledge vault.
Claim tracking integrates with a file-based vault and requires adherence to strict prose-as-title formatting and evidence provenance rules. No external dependencies are needed, but the vault environment must support structured claim documentation.
Documenting boundary conditions is necessary to define the parameter limits under which a scientific proposition remains valid, ensuring the generalizability of simulation analysis results is clearly bounded and evidence provenance is maintained.
Monitor documented claims through a status dashboard that provides an overview of all claim health and statuses. This allows researchers to track the lifecycle of scientific propositions and review accumulated evidence across simulation runs.
swarm-ai-safetyParse SWARM run metadata and summary files into structured vault notes.
vadirnAssign a 1-10 confidence score to decisions or claims with reasoning.
edonyzpcIdentify broken wikilinks, orphan notes, missing embeds, and inconsistent backlinks in Obsidian vaults.
shakedfbResearch topics and synthesize provenance-backed findings into a knowledge graph.
