plan-eng-review

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes various local binaries and shell commands related to the 'gstack' ecosystem (e.g., gstack-update-check, gstack-config, gstack-telemetry-log, gstack-review-log) and uses standard developer tools like git and codex for project analysis.
  • [EXTERNAL_DOWNLOADS]: The skill performs telemetry operations via gstack-telemetry-log and requests to open an external URL (https://garryslist.org/posts/boil-the-ocean) in the user's browser for educational purposes.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing and reviewing untrusted data from design documents and plans located in the project directory.
  • Ingestion points: Reads project-specific design documents (e.g., *-design-*.md) from ~/.gstack/projects/.
  • Boundary markers: Does not utilize explicit delimiters or 'ignore' instructions when interpolating the content of design documents into its review process.
  • Capability inventory: Possesses the ability to execute shell commands, invoke external AI tools (codex), and write various files (test plans, logs, and review summaries) to the local file system.
  • Sanitization: No validation or sanitization is performed on the ingested documentation content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 12:53 PM
Security Audit — agent-trust-hub — plan-eng-review