agent-identity-evolution

Pass

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: All external communications are directed to the vendor's official domain (agentpersona.live) to support the skill's identity evolution features.
  • [COMMAND_EXECUTION]: The skill utilizes standard system binaries curl and jq for performing API requests and parsing JSON data.
  • [EXTERNAL_DOWNLOADS]: Identity configurations and updated system prompts are fetched from the vendor's infrastructure as intended by the skill's design.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it aggregates content from sources like GitHub and Twitter into agent instructions. Ingestion points: GitHub, websites, RSS, Twitter, LinkedIn (SKILL.md). Boundary markers: None specified. Capability inventory: Limited to API interaction via curl and jq. Sanitization: Mitigated by a three-stage distillation pipeline and a required manual review step (/evolve review).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 25, 2026, 04:41 PM
Security Audit — agent-trust-hub — agent-identity-evolution