AgentDB Memory Patterns

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation instructs users to execute npx agentdb@latest, which downloads and runs the AgentDB CLI tool from the npm registry. This is a vendor-owned resource associated with the author (ruvnet).
  • [COMMAND_EXECUTION]: The skill provides instructions for several CLI operations, including database initialization (init), vector querying (query), and performance benchmarking (benchmark). It also includes commands for integrating AgentDB with Claude Code via the Model Context Protocol (MCP).
  • [PROMPT_INJECTION]: The skill provides patterns for storing and retrieving conversation history, creating a surface for indirect prompt injection if untrusted data is processed.
  • Ingestion points: Data enters the agent's context through adapter.insertPattern and db.storeMemory methods mentioned in SKILL.md.
  • Boundary markers: No specific boundary markers or 'ignore' instructions are provided in the code examples to isolate stored data from agent instructions.
  • Capability inventory: The skill utilizes local file-system access for database storage and CLI commands for querying. No dangerous system-level execution is triggered by stored data.
  • Sanitization: No explicit sanitization or content validation logic is demonstrated in the provided memory management examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 05:44 AM