The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from external sources including Figma frames, design components, product requirements, and task tickets. This creates a surface for indirect prompt injection where instructions could be embedded in the audited data. However, the skill implements a comprehensive 'Readiness Model' and 'Evidence Scale' (E0-E6) that requires the agent to verify claims against structural, documented, or runtime evidence rather than blindly trusting artifacts. This rigorous verification process, combined with the instruction to distinguish observed evidence from unknown data, acts as a significant control against the agent being misled by instructions found within the input data.

balise-handoff