The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an installation script hosted on the vendor's GitHub repository (https://raw.githubusercontent.com/tryopendata/opendata/main/scripts/install-cli.sh). This is used to install the opendata CLI tool required for authentication.
[COMMAND_EXECUTION]: The instructions suggest that the user can install the necessary CLI by piping the remote script to bash (curl -fsSL ... | bash). While this is a high-privilege execution pattern, it is directed to the user and targets the vendor's own verified repository.
[CREDENTIALS_UNSAFE]: The skill instructs the agent to search for API tokens in the user's home directory at ~/.config/opendata/auth.json. This is the standard path for the vendor's configuration and is necessary for authorized API access.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the api.tryopendata.ai endpoint.
Ingestion points: API responses in JSON, CSV, TSV, and XLSX formats (SKILL.md).
Boundary markers: None identified in the instructions for separating data from prompt instructions.
Capability inventory: The skill uses curl to perform network operations and provides methods for executing SQL queries on the server side.
Sanitization: No specific sanitization or filtering of the API-returned content is mentioned, which is common for data research tools.

opendata-api