The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes dashboard and folder definitions (titles, descriptions, tags, and queries) from external JSON files or CLI outputs.
Ingestion points: Dashboard and folder definitions are ingested via the oodle CLI commands dashboards get and folders get, and from local files supplied to the -f flag in create and update commands.
Boundary markers: The instructions do not define explicit delimiters or warnings to ignore embedded instructions within the dashboard JSON fields.
Capability inventory: The skill executes oodle CLI commands for creating, updating, and deleting resources, and uses jq and shell redirection for data manipulation and file writing.
Sanitization: No sanitization or validation of the text content within the JSON fields is performed before it is processed by the agent.
[COMMAND_EXECUTION]: The skill relies on executing shell commands, specifically the oodle CLI tool and jq for JSON parsing. These provide a vector for command execution based on the resource IDs and file paths handled by the agent.
[EXTERNAL_DOWNLOADS]: The skill instructs the installation of the oodle CLI tool via Homebrew (brew install oodle-ai/oodle/oodle). This is a legitimate installation of the author's own tooling according to the vendor context.

oodle-dashboards