The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses a Python utility, atoms-search.py, to query https://api.xyskill.xyz/v1/atoms/search. This endpoint is a vendor-owned resource belonging to the skill author (xyaz1313) and is used strictly to retrieve business methodology data. The script retrieves structured JSON data and does not execute remote code.
[INDIRECT_PROMPT_INJECTION]: The skill implements a Retrieval-Augmented Generation (RAG) pattern, fetching external 'atoms' from a JSONL file or a remote API to incorporate into its strategic analysis.
Ingestion points: Data is ingested from references/atoms.jsonl and https://api.xyskill.xyz/v1/atoms/search based on user-provided industry keywords.
Boundary markers: The skill uses explicit source IDs (e.g., XY-MA-1300) and instructions to treat the data as 'ammunition' (弹药) or 'evidence' (实证) for its conclusions, which provides a level of separation between retrieved data and system instructions.
Capability inventory: The skill's primary capabilities are text generation and data retrieval; the atoms-search.py script performs local file reads and HTTP POST requests but does not possess shell execution or file-write permissions.
Sanitization: While explicit sanitization of the retrieved text is not detailed in the prompt, the agent is instructed to use the atoms as specific references within its own established strategic framework.
[DATA_EXPOSURE]: The provided Python script attempts to read ~/.xy/config.json to locate the local database path. This is a standard configuration pattern for the vendor's tools, and the script does not transmit the contents of the configuration file or other sensitive user data over the network.

xy-playbook