niche-signal-discovery

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses a local, deterministic Python script (scripts/analyze_signals.py) to process data. The script relies entirely on standard libraries (csv, json, sys, re, argparse) and does not perform any network operations, dynamic code execution (eval/exec), or subprocess calls.
  • [SAFE]: Data enrichment is performed through the deepline CLI tool, which is a platform-integrated tool for batch enrichment. The workflow is transparent and requires explicit user commands for data fetching.
  • [SAFE]: No prompt injection or behavior override patterns were found in the instructions. All 'IMPORTANT' and 'CRITICAL' markers are used for legitimate operational guidance (e.g., deduplication and file verification).
  • [SAFE]: External service references (Apollo, Exa, Crustdata) are well-known B2B data providers and are consistent with the skill's purpose of signal discovery for sales and marketing teams.
  • [SAFE]: No obfuscation (Base64, hex encoding, or hidden characters) or persistence mechanisms were detected in the skill files or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 06:06 AM