icp-persona-builder

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes data from untrusted external websites. 1. Ingestion points: Untrusted content is retrieved via WebFetch and WebSearch during the research phase described in SKILL.md. 2. Boundary markers: The skill lacks explicit delimiters or instructions to ignore malicious commands embedded in the fetched web content. 3. Capability inventory: The skill can read local client context files, perform network-based research, and write persona assets to the local filesystem (SKILL.md). 4. Sanitization: No explicit sanitization or filtering of fetched web content is performed before processing.
  • [NO_CODE]: This skill consists entirely of instructional prompts and configuration in SKILL.md and does not include any executable code scripts or binaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 09:59 PM