dev-ship
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs shell operations using the
ghCLI for repository management andpnpmornpx nextfor project builds and linting. These are standard developer operations intended to facilitate the skill's primary purpose. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it processes external content from GitHub issues to drive its autonomous development cycles.
- Ingestion points: Fetches issue bodies and labels using
gh issue listandgh issue viewcommands inSKILL.md. - Boundary markers: Absent; the agent is instructed to read the entire issue body for acceptance criteria without explicit delimiters or instructions to ignore embedded adversarial content.
- Capability inventory: Includes the ability to write code to the file system, execute arbitrary project build scripts, and interact with the GitHub API for PR and issue management.
- Sanitization: Employs regex to validate numeric issue IDs, but does not sanitize the natural language body of the issues which serve as the implementation guide.
Audit Metadata