The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands via ssh to interact with the Docker daemon on a local device (raspberrypi.local) to identify running bot instances.
[COMMAND_EXECUTION]: The skill performs network requests using curl to various local API endpoints. It uses the -k flag to allow insecure connections (skipping SSL certificate validation), which is typical for local network environments.
[PROMPT_INJECTION]: The skill processes external data from command outputs and API responses, creating a surface for indirect prompt injection.
Ingestion points: Output from docker ps and responses from six distinct API endpoints (/api/health, /api/status, /api/signals/current, /api/trading, /api/alerts, /api/performance/summary).
Boundary markers: None present; data is synthesized directly into a narrative report.
Capability inventory: SSH command execution and local network access via curl.
Sanitization: No validation or escaping of the ingested data is performed before the agent processes the information.
[SAFE]: All identified command execution and network operations are strictly aligned with the documented purpose of the skill. The resources accessed are restricted to the local network (.local domain), and no hardcoded credentials or exfiltration patterns were found.

bot-status