create-rule

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes existing project instruction files (CLAUDE.md and .claude/rules/) which may contain untrusted data from the repository.\n
  • Ingestion points: Phase 4 of SKILL.md reads existing rule files from .claude/rules/ and CLAUDE.md into the agent context.\n
  • Boundary markers: Absent. The sub-agent prompts do not include instructions to isolate or ignore potentially malicious instructions embedded in the project files.\n
  • Capability inventory: The skill uses Write, Edit, and Bash(mkdir) tools to create or modify local configuration files based on the processed data.\n
  • Sanitization: Absent. The skill does not validate or sanitize the content extracted from existing rules before using it to generate and write new rule structures.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 08:42 PM
Security Audit — agent-trust-hub — create-rule