The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use bash and curl to interact with the Pipa gateway API. It explicitly directs the agent to check for environment variables like PIPA_API_BASE_URL and PIPA_EXECUTION_SECRET and execute shell commands to manage trigger subscriptions (create, list, delete).
[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill is designed to ingest data from external applications (GitHub, Linear, Slack, Asana) and incorporate it into trigger_prompt instructions for future execution.
Ingestion points: Processes untrusted event payloads from external provider integrations and user-defined request text.
Boundary markers: While the skill advises using "normalized event context," it does not specify the use of strict data delimiters or instructions to the LLM to ignore potentially malicious content within the external event data.
Capability inventory: The generated prompts are intended to trigger actions in external tools, including write operations like posting comments to Asana or summarizing GitHub PRs.
Sanitization: There is no evidence of strict schema validation or sanitization of the external event data before it is interpolated into the executable prompt strings.

pipa-triggers