The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data from the local project environment, specifically Salesforce metadata XML files and outputs from Salesforce CLI commands. This data is interpolated into the agent's context and could contain malicious instructions designed to manipulate the agent's behavior.\n- Ingestion points: Metadata configuration files (e.g., .namedCredential-meta.xml, .connectedApp-meta.xml, .object-meta.xml) read using Read, Glob, and Grep tools, as well as command outputs from Bash(sf *).\n- Boundary markers: No specific boundary markers or instructions to ignore embedded prompts are provided in the skill instructions.\n- Capability inventory: The agent possesses capabilities to Write and Edit project files, and execute Salesforce CLI commands (sf) via Bash to modify platform configuration.\n- Sanitization: The instructions do not define sanitization or validation logic for the content read from external files before processing.

sf-integration