The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data from the Linear API which may contain untrusted content.
Ingestion points: Data enters the agent context through the output of scripts/graphql, which returns JSON responses from https://api.linear.app/graphql.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the API response as untrusted data or to ignore embedded instructions within the response.
Capability inventory: The skill includes the ability to perform network requests via curl in scripts/graphql and execute Python scripts (scripts/check-response.py, scripts/read-credential.py).
Sanitization: scripts/check-response.py performs basic JSON structure validation and checks for GraphQL error flags, but does not sanitize the content of the data for potentially malicious natural language instructions.
[DATA_EXFILTRATION]: The skill includes a utility scripts/read-credential.py that can read any environment variable matching a specific naming pattern.
Evidence: The script scripts/read-credential.py takes a command-line argument and prints the corresponding environment variable value to standard output. While used by the scripts/graphql wrapper specifically for LINEAR_API_KEY, the standalone script provides a mechanism for the agent to access other sensitive environment variables (e.g., credentials for other services) if it chooses to invoke the script directly with a different variable name.

linear