external-pr-viability
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions define a legitimate and transparent analytical process using the official GitHub CLI (
gh) to query public repository metadata. The operations performed are read-only interactions with a well-known service (GitHub) to support implementation decisions. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, specifically GitHub PR metadata and repository documentation files like CONTRIBUTING.md.
- Ingestion points:
gh pr listJSON output,gh apimembership responses, and the contents ofCONTRIBUTING.mdandREADME.mdmentioned in SKILL.md. - Boundary markers: Not explicitly defined for markdown content, though PR data is processed as structured JSON.
- Capability inventory: The agent has the capability to execute
ghCLI commands and update local tracking records (file writes). - Sanitization: No specific input sanitization or explicit 'ignore instructions' delimiters are provided for the external text content, though the risk is minimal given the specific data-counting nature of the tasks.
Audit Metadata