agentic-workflows
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches workflow prompts and configuration files from the
github/gh-awrepository. - [INDIRECT_PROMPT_INJECTION]: The skill incorporates instructions from external and local sources, creating a potential surface for indirect prompt injection.
- Ingestion points: Remote markdown files from
github/gh-awand the local.github/aw/instructions.mdfile (relative to the user's repository). - Boundary markers: None; instructions are followed directly, and the local overlay is given precedence over default instructions.
- Capability inventory: The loaded prompts guide the agent in creating, updating, and debugging workflows, as well as managing tool mappings and cost optimization.
- Sanitization: None; the agent consumes the instruction content as provided by the file system or remote repository.
Audit Metadata