oc-deploy
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various CLI tools to manage the repository and deployment lifecycle.
- Evidence: Uses
gitfor status and tag management,gh(GitHub CLI) for workflow dispatch and log retrieval,pnpmfor package building and E2E testing, andgofor runtime verification. - [EXTERNAL_DOWNLOADS]: The skill interacts with external services to manage dependencies and trigger remote builds.
- Evidence: Triggers GitHub Action workflows using
gh workflow runand utilizespnpmwhich fetches external Node.js dependencies. - [INDIRECT_PROMPT_INJECTION]: The skill reads external, potentially untrusted data which could influence agent behavior if malicious content is present.
- Ingestion points: Processes GitHub Action logs via
gh run view --log-failedand repository metadata viagit status(SKILL.md, references/runbook.md). - Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard instructions embedded within the logs or external command outputs.
- Capability inventory: The skill can execute powerful commands including triggering remote workflows, pushing git tags, and running arbitrary build scripts (references/runbook.md).
- Sanitization: There is no evidence of sanitization or filtering of log content or external metadata before it is presented to the agent context.
Audit Metadata