spec-miner

Warn

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: MEDIUMPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted source code from the local environment.
  • Ingestion points: Source files are read using the Read, Grep, and Glob tools.
  • Boundary markers: No delimiters or explicit instructions are provided to the agent to ignore embedded commands within the code.
  • Capability inventory: The skill utilizes Bash and general file system read/write operations.
  • Sanitization: There is no validation or sanitization of code content before it is added to the agent's context.
  • [DATA_EXFILTRATION]: The skill instructions direct the agent to locate and read sensitive configuration files (e.g., .env) and authentication logic to facilitate mapping the system architecture, which constitutes high-severity data exposure.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 24, 2026, 01:36 AM
Security Audit — agent-trust-hub — spec-miner