execplan-improve

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from the codebase to generate and rewrite execution plans.
  • Ingestion points: The skill reads .agent/execplan-pending.md, .agent/PLANS.md, and any source code files referenced or adjacent to the plan as described in Steps 1, 2, and 3 of SKILL.md.
  • Boundary markers: No explicit delimiters or boundary instructions are utilized to differentiate the analyzed code content from the skill's own operational logic.
  • Capability inventory: The skill possesses the capability to perform a file-read on arbitrary paths within the repository and worktree, and a file-write to rewrite the execution plan file as described in Step 5 of SKILL.md.
  • Sanitization: There is no evidence of sanitization, validation, or filtering of the content extracted from external files before it is incorporated into the rewritten execution plan.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 02:58 PM
Security Audit — agent-trust-hub — execplan-improve