Skills
Navigate
skills/aiskillstore/marketplace/agentdb-learning-plugins/Gen Agent Trust Hub

agentdb-learning-plugins

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Documentation provides examples for managing learning plugins via the agentdb CLI tool using npx.
  • [EXTERNAL_DOWNLOADS]: The skill references external project repositories and documentation websites for the agentic-flow and agentdb ecosystems.
  • [PROMPT_INJECTION]: The skill describes a pattern for ingesting untrusted agent experience data to build context for reasoning, creating an indirect prompt injection surface.
  • Ingestion points: adapter.insertPattern (SKILL.md).
  • Boundary markers: Absent in the documentation examples.
  • Capability inventory: Includes npx agentdb CLI operations, model training (adapter.train), and context synthesis (adapter.retrieveWithReasoning).
  • Sanitization: Not demonstrated in the simplified documentation examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 04:42 AM
Security Audit — agent-trust-hub — agentdb-learning-plugins