deployment-cicd
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [CREDENTIALS_UNSAFE]: The Docker Compose example in
SKILL.mdincludes a hardcoded placeholder password (POSTGRES_PASSWORD: password). While a standard practice for local development documentation, it matches credential exposure patterns. - [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to instructions embedded in project configuration files it processes.
- Ingestion points: The skill is triggered by content in
Dockerfileand GitHub Actions workflow files (.github/workflows/*.yml). - Boundary markers: The instructions lack delimiters or explicit warnings for the agent to treat the content of these files as untrusted data.
- Capability inventory: The skill facilitates the execution of powerful build and deployment tools such as
npm,docker,vercel, andflybased on project content. - Sanitization: There are no defined processes to sanitize or validate the instructions found within analyzed project files.
- [COMMAND_EXECUTION]: The skill provides numerous templates that involve executing system commands for software building, testing, and deployment (e.g.,
npm run build,docker compose up,vercel rollback). - [EXTERNAL_DOWNLOADS]: The CI/CD recipes reference and download GitHub Actions from established organizations such as
actions/anddocker/, which are well-known and trusted sources.
Audit Metadata