The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of Bash to execute local and remote commands via SSH, including Docker management operations and connectivity tests.
Evidence includes use of ssh, docker ps, docker exec, and nc -z to manage services on infra.local and other hosts.
[EXTERNAL_DOWNLOADS]: The skill uses curl to perform health checks on internal network services such as Langfuse and Home Assistant.
Evidence: curl -s -o /dev/null -w "%{http_code}" http://infra.local:3000.
[DATA_EXPOSURE_AND_EXFILTRATION]: Documentation provides guidance on where local secrets and SSH configurations are stored to facilitate management.
Evidence: References to ~/.ssh/config, ~/.zshrc, and .env files for managing connection aliases and environment variables.
[PERSISTENCE_MECHANISMS]: The skill includes a workflow for configuring passwordless SSH access by appending aliases and configuration to the user's SSH config file.
Evidence: Use of cat >> ~/.ssh/config to automate the setup of connection aliases for infrastructure hosts.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes output from remote logs and service status commands.
Ingestion points: Remote command output from ssh (e.g., docker logs, docker ps), and curl HTTP responses.
Boundary markers: No explicit delimiters or boundary markers are used when processing remote service data.
Capability inventory: The skill utilizes bash, ssh, and network tools to interact with remote environments.
Sanitization: No explicit sanitization or validation of the ingested remote data is performed before it is presented to the agent.

infra-manage-ssh-services