slfg

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-supplied input through the $ARGUMENTS variable in the /ce:plan command without explicit delimiters or sanitization instructions. This creates a surface for indirect prompt injection where a malicious feature description could attempt to influence the agent's behavior during the automated workflow.
  • Ingestion points: SKILL.md (via $ARGUMENTS)
  • Boundary markers: Absent
  • Capability inventory: The workflow includes automated code generation (/ce:work), browser-based testing (/compound-engineering:test-browser), and direct file system modification via autofixing (/ce:review mode:autofix).
  • Sanitization: No input validation or instruction filtering is present in the skill definition.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 04:47 PM