The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses kubectl and curl to manage deployments and verify service status. These operations are restricted to the allowed-tools configuration and align with the skill's purpose as a production checklist.
[REMOTE_CODE_EXECUTION]: Automated detectors identified piped execution patterns involving curl and python3. A manual review of SKILL.md reveals that these patterns are used strictly for parsing and formatting JSON data from trusted sources (the Bright Data API and local deployment health checks) using python3 -c for parsing and python3 -m json.tool for pretty-printing. They do not execute external scripts or arbitrary code.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from Bright Data's API and local health endpoints during status verification.
Ingestion points: curl calls in SKILL.md targeting https://api.brightdata.com/zone/get_active_zones, api/v2/status.json, and http://localhost:8080/health.
Boundary markers: None implemented in the processing logic.
Capability inventory: kubectl for infrastructure management and curl for network requests.
Sanitization: Standard JSON parsing is used; while it doesn't provide instruction filtering, the inputs are structured status objects from the user's own infrastructure or service provider, posing minimal risk.

brightdata-prod-checklist