skills/cpfiffer/central/agent-profile/Gen Agent Trust Hub

agent-profile

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (LOW): The skill facilitates querying agent profiles (names, descriptions) from the AT Protocol, which are attacker-controllable sources. This creates a surface where an agent could be influenced by malicious instructions embedded in a profile. * Ingestion points: Agent profile fields (name, description) ingested via tools.registry in SKILL.md. * Boundary markers: Delimiters or 'ignore instructions' markers are absent when processing external profiles. * Capability inventory: The skill uses tools.registry which performs network operations on the AT Protocol. * Sanitization: No sanitization or verification of the external metadata is defined.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:47 PM