The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard system commands such as whoami, hostname, and otelcol-contrib validate to manage authentication status and verify configuration integrity before deployment.
[PRIVILEGE_ESCALATION]: The configuration reference for Docker monitoring describes mounting the Docker socket (/var/run/docker.sock) and running the collector container as root (user: "0:0"). The skill appropriately warns the user that this provides root-level control over the host system.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface by reading local project configuration files to identify services to monitor.
Ingestion points: SKILL.md (Step 2) instructs the agent to read docker-compose.yml, Dockerfile, pyproject.toml, and package.json to identify databases, caches, and queues.
Boundary markers: Absent for these file-reading operations.
Capability inventory: The skill can execute shell commands for validation and uses a project-scoped write token for telemetry export.
Sanitization: No specific sanitization or validation of the contents of these configuration files is described beyond structure validation via the OTel binary.

logfire-infrastructure