The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection because it reads and processes data from external files that could contain untrusted input.
Ingestion points: Reads external configuration and component definitions from docs/component-selections.md, docs/design-constraints.json, and reference materials in datasheets/ as outlined in SKILL.md.
Boundary markers: No specific delimiters or boundary isolation instructions are provided to prevent the model from executing text-based commands embedded within these files.
Capability inventory: The skill utilizes the Write tool and multiple mcp__kicad-sch__* tools to modify project files such as hardware/*.kicad_sch and docs/schematic-status.md.
Sanitization: Lacks explicit validation or parsing sanitization to separate structural parameters from executable or instructional text fields.

eda-schematics