kohls-research

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a bash script (scripts/crawlora.sh) to interface with the Crawlora API. The script is well-structured, using set -euo pipefail for robustness and utilizing curl with proper argument handling (--data-urlencode) to prevent shell injection during query parameter construction.
  • [CREDENTIALS_UNSAFE]: The skill correctly instructs users to manage their API keys via an environment variable (CRAWLORA_API_KEY) and includes explicit warnings against hardcoding or committing secrets.
  • [EXTERNAL_DOWNLOADS]: The skill communicates exclusively with api.crawlora.net, which is the official infrastructure for the skill's author (crawlora-org). No unauthorized third-party downloads or remote code executions were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 06:09 PM
Security Audit — agent-trust-hub — kohls-research