giga-workflow
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the execution of local automation scripts for workspace management. Evidence includes the use of the
bunruntime to execute scripts such assrc/giga/giga-watch.ts,src/giga/run-submodule.ts, andsrc/giga/commit-propagator.ts. - [DYNAMIC_EXECUTION]: The skill supports the runtime execution of TypeScript files and the invocation of external tools via the user-configurable
PANTHEON_CLIenvironment variable. This allows the agent to run a commit message generator path provided in the environment. - [INDIRECT_PROMPT_INJECTION]: The skill identifies a potential surface where untrusted data from submodule changes could influence the generated commit messages. 1. Ingestion points: File changes within the
orgs/directory. 2. Boundary markers: None specified in the documentation for separating ingested content from instructions. 3. Capability inventory: Execution of local scripts and git operations including commit and tagging. 4. Sanitization: No explicit sanitization or filtering logic is described for the data processed by the message generator.
Audit Metadata