openclaw-release-plugin-testing
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use several powerful CLI tools, including
pnpm,git, and the GitHub CLI (gh). These are used to query repository state, execute local Docker-based tests, and trigger remote CI/CD workflows viagh workflow run. - [INDIRECT_PROMPT_INJECTION]: The skill operates on data retrieved from the software repository and system environment, which could potentially contain malicious instructions if the repository is compromised.
- Ingestion points: The agent reads data from
git status,pnpm changed:lanes --json,plugins list --json, andopenclaw status --json(SKILL.source.md). - Boundary markers: The instructions do not define clear delimiters or use "ignore embedded instructions" patterns when processing the output from these commands.
- Capability inventory: The agent is authorized to execute package management commands (
pnpm), source control operations (git), and workflow orchestration (gh) in both local and remote (GitHub Actions) environments. - Sanitization: There are no explicit instructions for sanitizing or validating the JSON/text outputs received from the project's tools before the agent processes them.
Audit Metadata