The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill processes external data from files and the web which could contain malicious instructions designed to influence the agent's output.
Ingestion points: Reads files from paths provided in vision_input and existing_okrs, and fetches external web content via the autonomous-research mixin (Phase 2 and Phase 3).
Boundary markers: The instructions do not define any delimiters or system warnings to the LLM to ignore potentially malicious instructions embedded in the external content.
Capability inventory: The skill possesses capabilities for file system reads (via the input paths), web searching/fetching, and command execution for diagram rendering.
Sanitization: No validation or sanitization logic is specified for the data imported from external strategy documents or web sources.
[COMMAND_EXECUTION]: Diagram Rendering. The skill uses the @mermaid-js/mermaid-cli tool (mmdc) to render OKR alignment trees and performance dashboards. This is a standard and expected behavior for skills utilizing the diagram-rendering mixin.

okr-definition