write-rule
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection. Malicious data within the agent's context could influence the 'intent gathering' phase, leading the agent to generate rules that 'allow' dangerous operations or disable security guardrails. (Ingestion points: Step 1
- intent gathering in SKILL.md; Boundary markers: Absent; Capability inventory: Writing to global and local JS configuration files, execution of commands via npx; Sanitization: Absent).
- [COMMAND_EXECUTION]: The skill generates and writes Javascript configuration files (
nyolo.config.js) to both project-specific and global directories (~/.claude/). These files contain executable logic and functional callbacks that are interpreted at runtime by the nyolo security hook. - [EXTERNAL_DOWNLOADS]: The skill instructs the use of
npx nyolo, which downloads and executes the nyolo package from the npm registry, and uses ESM imports for the package in generated code.
Audit Metadata