The Agent Skills Directory

[SAFE]: No security issues were identified in the skill instructions or reference files. The documentation for API tokens and headers is provided for informational purposes within the context of Celigo's official services. The skill references legitimate platform domains and includes installation instructions for related tools from the same verified vendor. The use of a local 'learnings' file to track platform knowledge is implemented as a standard feature for improving the assistant's performance over time without introducing unsafe command execution or data exfiltration risks.
[PROMPT_INJECTION]: The skill implements a mechanism to persist insights in 'references/learnings.md' to maintain continuity. * Ingestion points: Information entered by the user or extracted from platform references is documented in the learnings file. * Boundary markers: No specific delimiters or safety instructions are defined for the learning file content. * Capability inventory: The skill is restricted to file read/write operations and natural language responses. * Sanitization: Data is appended based on session discoveries without explicit sanitization. This implementation follows standard platform patterns for persistent knowledge storage and is considered safe for the intended use case.

sales-celigo