reverse-prompt

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to generate structured execution briefs that can trigger subsequent actions.
  • Ingestion points: The workflow in SKILL.md involves extracting intent, scope, and constraints from raw user-provided "rough requests."
  • Boundary markers: The instructions do not specify the use of delimiters or explicit boundary markers to isolate the untrusted user input from the brief templates defined in references/brief-template.md.
  • Capability inventory: The skill is designed to drive subsequent research, plan, or implement phases (as described in references/decision-rules.md) which typically involve file system access and shell command execution.
  • Sanitization: There are no instructions to sanitize, escape, or validate the user's input before it is restructured into the rewritten prompt, allowing potential malicious instructions to persist into the execution phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 05:29 PM