verification
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies and executes shell commands inferred from the local workspace to launch applications and run test suites. It parses files such as
Makefile,package.json,docker-compose.yml, and CI configurations to determine these commands. - [DYNAMIC_EXECUTION]: The skill generates and runs temporary automation scripts (e.g., using Playwright via
npx) and injects JavaScript probes into the browser to measure layout shifts and detect dead UI controls during live verification. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository to influence its execution flow, creating a potential vulnerability surface.
- Ingestion points: The skill reads repository manifests, CI configuration,
Makefile,docker-compose.yml, anddocs/verification.md(as described inSKILL.mdandreference.md). - Boundary markers: No explicit markers are used to distinguish between trusted instructions and data retrieved from the repository.
- Capability inventory: The skill can execute arbitrary shell commands (test suites, launch scripts) and perform network/database operations (
curl, database clients) as part of the verification process. - Sanitization: There is no mention of sanitizing or validating the commands inferred from the repository files before execution.
Audit Metadata