The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run in scripts/airtap.py to interact with an external CLI tool named openclaw. This is used to relay Airtap task updates to the OpenClaw service. The implementation follows security best practices by using a list of arguments to avoid shell injection and verifying that the target binary name is exactly 'openclaw'.
[INDIRECT_PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests and processes task progress updates from the Airtap API.
Ingestion points: Task details and agent messages are retrieved via the api_get_task_details function in scripts/airtap_common.py.
Boundary markers: None identified in the prompt templates or processing logic.
Capability inventory: The skill can perform network requests to the Airtap API using requests and execute the openclaw binary via subprocess.run in scripts/airtap.py.
Sanitization: The skill performs basic text normalization and truncation but does not implement specific filtering to prevent embedded instructions in task updates from influencing the agent's behavior.

airtap