verification-before-completion
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform shell-based operations including Git commands (
git diff,git commit) to manage task lifecycle and state. Additionally,expanded-closeout.mdspecifies the execution of a Python utility (<aegis-workspace-helper>) to bundle and check workspace integrity. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted data such as source code diffs, test evidence, requirement documents, and architecture baselines.
- Ingestion points: Reads
TaskStartSnapshot, code diffs, test output, and externalui-ux-governancedocumentation. - Boundary markers: While the skill uses an
<EXPLICIT-MODE-GATE>for its own activation, it lacks explicit delimiters or sanitization instructions for the content of the evidence it processes. - Capability inventory: The skill has the authority to write to the file system (Git commits) and execute local Python scripts.
- Sanitization: There are no explicit instructions to sanitize or escape the content of the external evidence before it is incorporated into the agent's context.
Audit Metadata