Skills
Navigate
skills/full-stack-skills/stitch-skills/openspec-explore/Gen Agent Trust Hub

openspec-explore

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local codebase, specifically configuration files and existing project artifacts, to inform the agent's reasoning.
  • Ingestion points: Reads openspec/config.yaml (or .yml) and existing artifact files located in artifactPaths returned by the CLI.
  • Boundary markers: Explicitly instructs the agent: 'do NOT copy them [context/rules] into the conversation or into any artifact you create.'
  • Capability inventory: Executes openspec CLI commands via bash and writes markdown artifacts to the local filesystem.
  • Sanitization: Relies on specific instructions for the agent to evaluate conditions against explored changes and use provided templates rather than raw external content.
  • [COMMAND_EXECUTION]: The skill utilizes the openspec CLI tool to perform operations such as listing stores, checking change status, and generating new artifacts.
  • The execution is restricted to the openspec command as defined in the allowed-tools frontmatter.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 07:51 AM