The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute oodle, curl, and jq commands. These are standard operations for managing the service, confirming reachability, and processing JSON output.
[EXTERNAL_DOWNLOADS]: The skill documents the installation of the vendor's official CLI tool via Homebrew (brew install oodle-ai/oodle/oodle). This is an expected installation path for the service provided by the skill author.
[INDIRECT_PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it reads monitor configurations from the service (oodle synthetic-monitors get) and processes them.
Ingestion points: Monitor configuration data is retrieved from the Oodle API and stored in monitor.json or monitor.new.json (SKILL.md).
Boundary markers: None present; data is processed as standard JSON.
Capability inventory: The skill uses subprocess calls to oodle, curl, and jq (SKILL.md).
Sanitization: None specified; the skill relies on standard JSON parsing and CLI argument handling.
Assessment: This surface is typical for administrative monitoring tools and is considered safe given the intended primary purpose of the skill.

oodle-synthetic