narrative
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool for configuration resolution using 'jq' and for executing a local Python utility script ('bridge-citations.py'). These commands are part of the core narrative generation workflow and are restricted to processing local project files.
- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests untrusted markdown data from several subdirectories (e.g., '04-findings', '11-trends') without explicit boundary markers or sanitization, and possesses capabilities like the 'Bash' tool. This risk is minimized by the skill's focus on structured narrative transformation and evidence verification.
- [SAFE]: A thorough review of the 72 files found no hardcoded credentials, sensitive file access patterns (e.g., '.ssh', '.env'), or suspicious network activity. All external references are limited to the vendor's internal ecosystem or informational placeholders.
Audit Metadata