The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local Python script scripts/atoms-search.py to retrieve reference data. This script is invoked via the shell with arguments derived from user queries, which represents a standard operational capability but carries typical command execution risks if inputs are not properly handled by the execution environment.- [EXTERNAL_DOWNLOADS]: The scripts/atoms-search.py script performs network requests to https://api.xyskill.xyz/v1/atoms/search. This domain is a vendor-managed resource for the 'xy' skill suite, used to fetch the knowledge 'atoms' necessary for the skill's primary function.- [INDIRECT_PROMPT_INJECTION]: The skill processes data fetched from an external API or local JSONL files, creating a vulnerability surface where malicious external content could influence agent behavior.
Ingestion points: Data retrieved by atoms-search.py from the external api.xyskill.xyz endpoint or local atoms.jsonl files.
Boundary markers: The instructions do not define strict delimiters (like XML tags) or 'ignore' instructions for the retrieved content when it is integrated into the agent's context.
Capability inventory: The skill has the capability to execute a Python script and perform network operations through that script.
Sanitization: There is no explicit sanitization logic mentioned to filter or escape the content returned by the external API before the agent processes it.

xy-mode