skill-staleness-check
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs automated repository management by executing command-line utilities. It uses the GitHub CLI (
gh) to retrieve source code diffs and to handle branch management and pull request creation. It also executes a project-specific tool,intent stale, to programmatically check for version drift and metadata inconsistencies between the skill files and the published libraries. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could be leveraged to influence agent behavior through indirect prompt injection.
- Ingestion points: The skill ingests untrusted data from two primary external sources: webhook payloads containing package metadata and raw source code diffs fetched from upstream GitHub repositories as described in SKILL.md.
- Boundary markers: There are no explicit instructions or delimiters defined to isolate the processed source code diffs from the agent's internal instructions during the surgical regeneration of skill content.
- Capability inventory: The agent has high autonomy and write capabilities, including modifying local markdown (
SKILL.md) and JSON (sync-state.json) files, creating new git branches, and submitting code changes via pull requests. - Sanitization: The instructions do not describe any sanitization, validation, or filtering of the external source code content before it is used to drive the text generation process.
Audit Metadata