The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands such as curl and jq to interact directly with the Pipefy GraphQL API when standard toolkit operations fail.\n- [EXTERNAL_DOWNLOADS]: Data and authentication tokens are retrieved from official Pipefy domains app.pipefy.com and api.pipefy.com. These are recognized vendor resources for this skill.\n- [REMOTE_CODE_EXECUTION]: The skill includes a pattern where the output of a curl request is piped to python3 -c for processing. While this matches a remote execution pattern, the source is a trusted vendor-controlled authentication endpoint.\n- [DYNAMIC_EXECUTION]: A Python one-liner is used at runtime to parse JSON responses and extract authentication tokens. This script is static and performs standard data processing.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external API responses, which represents an ingestion surface. The risk is minimized by the use of structured JSON parsing tools.\n
Ingestion points: JSON responses from https://api.pipefy.com/graphql and https://app.pipefy.com/graphql as documented in SKILL.md.\n
Boundary markers: The skill relies on structured JSON formatting; no specific natural language delimiters or "ignore instructions" markers are mentioned for the response content.\n
Capability inventory: The agent has access to curl, jq, and python3 to process and potentially act on the ingested data.\n
Sanitization: No specific input sanitization is implemented for the retrieved content, though the skill focuses on verifying error codes within the structured data.

pipefy-api-fallback