The Agent Skills Directory

[SAFE]: The skill implements a robust quality assurance framework for the final phase of a software analysis workflow. It mandates a comprehensive self-validation checklist that covers story understanding, API analysis, design context, and component hierarchy, effectively preventing the production of incomplete or unverified implementation plans.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, potentially untrusted sources including Jira stories, Figma design metadata, and API specifications, which constitutes a theoretical attack surface for indirect prompt injection.
Ingestion points: Technical details are extracted from Jira 'Developer Notes', Figma design nodes, and Sitecore rendering contracts as defined in SKILL.md.
Boundary markers: The skill enforces the use of strict markdown templates and structured tables, providing a clear schema for the output and isolating external data into specific fields.
Capability inventory: The skill is restricted to file system operations within a designated path (.SS_WF/Agent/Analysis/) using the write_file tool. It does not have access to network or shell execution tools in this phase.
Sanitization: The instructions mandate strict adherence to 'Prop-Driven' models and 'Responsibility Matrices', ensuring that external labels and values are treated as data properties rather than instructions that could influence the agent's core logic or downstream execution.

analysis-output-contract