generate-execution-prompt

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill constructs an execution prompt by interpolating content from external files like plan.md and LESSONS.md into a pre-defined template.
  • Ingestion points: The generator reads plan.md, spec.md, and LESSONS.md from the run directory.
  • Boundary markers: Data is placed into defined {{slot}} regions within a markdown text fence.
  • Capability inventory: The generated output includes instructions for an agent to execute shell commands (e.g., build, test, lint) and perform file system operations.
  • Sanitization: The process incorporates a mandatory "Fresh-eyes check" where a subagent validates the generated prompt's viability before the user executes it.
  • [EXTERNAL_DOWNLOADS]: The skill references documentation and engineering guidelines from Anthropic's official platform and blog.
  • [COMMAND_EXECUTION]: The skill generates instructions for executing repository-specific build and test commands as part of its verification discipline.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 07:41 AM
Security Audit — agent-trust-hub — generate-execution-prompt