release-checklist
Warn
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: In Phase 2, the skill is instructed to detect and run every quality gate command discovered in the project via
package.jsonscripts,Makefile, CI configuration, orpyproject.toml. This dynamic discovery and execution of shell commands from local configuration files creates a security risk if the files contain malicious or unexpected commands. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external, potentially untrusted project data to make critical release decisions.
- Ingestion points: Reads content from
plans/<name>/tasks.md,plans/<name>/prd.md, and project configuration files (package.json,Makefile,pyproject.toml). - Boundary markers: Absent. The skill does not use delimiters or instructions to ignore potential injection attacks within the plan files.
- Capability inventory: The skill can execute shell commands through detected quality gates and perform file system writes to create
CHANGELOG.mdandrelease-checklist.md. - Sanitization: Absent. There is no validation or sanitization logic for the commands extracted from configuration files before they are passed to the shell.
Audit Metadata