skill-update-manager
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
scripts/check-updates.pyscript utilizes the GitHub CLI (gh) viasubprocess.runto query repository metadata and download source tarballs. This is a functional requirement for checking third-party skill updates. - [EXTERNAL_DOWNLOADS]: The skill fetches repository tarballs from GitHub for comparison purposes. These downloads target a well-known service and are handled using security best practices, such as applying the
tarfilemodule's 'data' filter during extraction to prevent path traversal attacks. - [INDIRECT_PROMPT_INJECTION]: As an update manager, the skill ingests data from external sources (GitHub repository content) and local configuration files (
.skill-lock.json). This creates a surface for indirect prompt injection if a source repository were compromised, though the skill operates primarily as a reporting tool to inform the user's next manual steps.
Audit Metadata