github-filing-registry

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Uses a shell dispatcher to execute subcommands using bash, jq, and the gh CLI for registry management and API interaction.- [EXTERNAL_DOWNLOADS]: Synchronizes registry state by fetching issue and pull request metadata from GitHub's official API.- [INDIRECT_PROMPT_INJECTION]: Processes and displays content from external repositories which could contain instructions intended to influence the agent.
  • Ingestion points: Retrieves titles and status information for issues and pull requests from the GitHub API via the gh CLI in the filing-registry script.
  • Boundary markers: The script uses distinct header blocks (e.g., === TRIAGE ... ===, === ACTIVITY LOG ... ===) to separate registry data from other agent output.
  • Capability inventory: The skill can perform local file read/write operations on the registry and execute GitHub API queries.
  • Sanitization: Uses jq to ensure valid JSON and string formatting, though it does not explicitly filter for prompt injection sequences within retrieved metadata titles.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 11:31 PM