topic-resolver

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted user queries to generate entities used by downstream search tools.
  • Ingestion points: User-provided topic strings processed in the 'The Problem' and 'Example Workflow' sections.
  • Boundary markers: No markers or explicit instructions are provided to the agent to ignore embedded commands within the user topic.
  • Capability inventory: The skill is designed to interface with 'oracle' and 'harvest' agents which possess network search and deep crawling capabilities.
  • Sanitization: The instructions lack any input validation, escaping, or filtering of the user-provided content before it is used to determine search targets.
  • [DATA_EXFILTRATION]: The skill instructions specify that topic resolutions are cached in the local file ~/.claude/topic-cache.jsonl. While intended for performance, access to the user's home directory configuration folders represents a surface for tracking or information exposure if the integrated agents are compromised.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 07:46 AM
Security Audit — agent-trust-hub — topic-resolver