The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized network operations were detected in the skill instructions.
[COMMAND_EXECUTION]: The skill instructions direct the agent to write an artifact named phase-1-dimensions.md to the local directory. This is a legitimate part of the skill's documented workflow for surfacing prompt dimensions.
[DATA_EXPOSURE]: The skill reads data from grain-context.md and various files in the references/ directory. This data access is restricted to the local file system and is necessary for the skill's stated purpose of extracting domain expertise requirements from a project context.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (referred to as a 'grain'). While it includes instructions to quote the grain verbatim in its output, there is an inherent risk that the input could contain instructions designed to influence downstream agents that process the generated phase-1-dimensions.md file. This is an architectural consideration for systems using this skill.
Ingestion points: Reads the 'grain' from grain-context.md or direct user arguments.
Boundary markers: The skill instructs the agent to quote the grain within a markdown block (> <quoted grain>).
Capability inventory: Uses Read, Write, Edit, Glob, Grep, and Agent tools for local file management and agent orchestration.
Sanitization: No specific sanitization or escaping logic is defined for the external content before it is written to the output file.

dimension-surfacing