The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external artifacts which may contain malicious instructions designed to manipulate the agent's meta-analysis or findings.
Ingestion points: Project artifacts read using the Read tool during the Structural Analysis phase.
Boundary markers: None defined; the skill processes the entire content of the provided artifact without isolation.
Capability inventory: The skill uses Read and Write tools to perform analysis and update the .prism-history.md log file.
Sanitization: No explicit sanitization or filtering of the artifact content is performed prior to analysis.

prism-reflect