release
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the GitHub CLI (
gh) to dispatch workflows. This command uses variables derived from external CI runs (candidate_run_id,candidate_attempt) in thePublish the retained bytessection of SKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external GitHub Actions runs to determine release readiness. 1. Ingestion points: Metadata, job conclusions, and artifact digests are read from
release-candidate.yml(SKILL.md). 2. Boundary markers: Absent; no delimiters distinguish untrusted CI data from instructions. 3. Capability inventory: Shell command execution viagh workflow run(SKILL.md). 4. Sanitization: Absent; the skill relies on human verification of the presented metadata before execution. - [EXTERNAL_DOWNLOADS]: The skill references resources from GitHub and NuGet for verification and publication. It also checks for updates in the
richlander/dotnet-skillsrepository, which is a vendor-owned resource. These are well-known or vendor services used for development workflows.
Audit Metadata