The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes curl for API interaction and python3 for processing JSON responses and encoding data. These commands are executed via the agent's terminal tool.
[DYNAMIC_EXECUTION]: The skill uses python3 -c to execute small, dynamically generated code snippets for URL-encoding search formulas.
[INDIRECT_PROMPT_INJECTION]: The skill fetches data from external Airtable bases, creating a vector for instructions to enter the agent's context.
Ingestion points: External data is ingested through curl responses from https://api.airtable.com containing record fields and table schemas.
Boundary markers: The skill does not provide specific delimiters or instructions to treat data retrieved from the API as potentially untrusted.
Capability inventory: The agent utilizes the terminal tool to run shell commands like curl and python3.
Sanitization: While the skill uses URL-encoding for outbound requests, it lacks sanitization or filtering for data retrieved from record fields before it is processed.

airtable