The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses sf org display and sf api request rest to interact with a connected Salesforce organization. These are standard operations for Salesforce developers and administrators using the Salesforce CLI.
[DATA_EXPOSURE_&_EXFILTRATION]: The skill fetches 'agentic setup categories' and 'prompts' from the Salesforce Connect API. This is metadata used to configure AI agents within the Salesforce ecosystem. The data is retrieved from the user's own authenticated org and presented back to the user within the agent session. No data is sent to external or unauthorized third-party domains.
[INDIRECT_PROMPT_INJECTION]: The skill processes JSON data from a Salesforce API. While the text field in the API response contains prompt strings, these are treated as data to be displayed rather than instructions to be executed by the agent. The skill includes clear documentation of the expected schema in references/api-response-schema.md, which helps the agent maintain context boundaries.
[SAFE]: The skill is authored by 'forcedotcom' (Salesforce) and interacts exclusively with Salesforce infrastructure (/services/data/...) using official tooling (sf CLI). The functionality aligns exactly with the stated purpose of fetching setup categories.

platform-agentsetup-categories-fetch