skills/mocasus/paleo/paleo-converse/Gen Agent Trust Hub

paleo-converse

Pass

Audited by Gen Agent Trust Hub on Jul 26, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill consists entirely of markdown instructions and metadata. No scripts, executables, or package dependencies are provided or referenced.
  • [PROMPT_INJECTION]: The skill defines rules for processing conversation history, which is a surface for indirect prompt injection.
  • Ingestion points: Conversation history turns (referred to in SKILL.md rules).
  • Boundary markers: Absent.
  • Capability inventory: None; the skill does not perform command execution, file writes, or network requests.
  • Sanitization: Absent.
  • [SAFE]: No obfuscation, persistence mechanisms, or unauthorized data access patterns were detected during the analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 26, 2026, 12:11 AM
Security Audit — agent-trust-hub — paleo-converse