agentic-workflows
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill retrieves Markdown instruction files and prompt templates from the 'github/gh-aw' repository. These files define the agent's behavior for creating, debugging, and managing workflows. Because these resources are hosted in an official repository by the service provider, this behavior is considered a standard functional requirement.
- [INDIRECT_PROMPT_INJECTION]: The skill has a defined surface for indirect instructions by loading remote files and local repository-specific overrides.
- Ingestion points: Remote workflow files from 'github/gh-aw' and the local file '.github/aw/instructions.md'.
- Boundary markers: The skill does not implement explicit boundary markers or safety instructions to ignore embedded commands, instead instructing the agent to follow the loaded files directly.
- Capability inventory: The instructions guide the agent in performing administrative and development tasks, including workflow design, creation, and debugging.
- Sanitization: There is no evidence of sanitization or validation for the content of the ingested Markdown files.
Audit Metadata