gtm-fix-guide
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill acts as a pure data transformer for local GTM debugging files. It does not attempt to access sensitive system files (e.g., SSH keys, credentials), nor does it initiate network requests or execute shell commands. All operations are confined to reading and writing text-based files within the project root.- [INDIRECT_PROMPT_INJECTION]: The skill exposes a potential surface for indirect prompt injection as it ingests data from external JSON and Markdown files, though the risk is negligible due to the absence of high-privilege capabilities.
- Ingestion points: The skill reads
gtm-test-results.jsonandgtm-context.mdto inform its logic. - Boundary markers: There are no explicit markers or specialized instructions used to prevent the agent from following directions that might be embedded within the source files.
- Capability inventory: The skill is only permitted to read and write local files. It has no capabilities for network communication, system command execution (RCE), or credential access.
- Sanitization: The skill performs no sanitization or validation on the input data before including it in the generated markdown guide, leaving the possibility for the output document to be influenced by untrusted input content.
Audit Metadata