The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user-provided briefs, research artifacts, and existing codebases to generate its output.
Ingestion points: Processes data from .jez/artifacts/research-brief-{topic}.md, CLAUDE.md, and arbitrary project files.
Boundary markers: Absent; the instructions do not specify delimiters or warnings for the agent when processing these external files.
Capability inventory: Includes Read, Write, Edit, Glob, Grep, and Bash for file system operations.
Sanitization: Not explicitly implemented for the ingested content.
[DATA_EXPOSURE]: The skill includes instructions to search ~/Documents/ to identify patterns in the user's existing technology stack. While this involves broad filesystem access, it is contextually relevant to the skill's purpose of planning a compatible tech stack and does not involve external data transmission.

roadmap