The Agent Skills Directory

[SAFE]: The skill operates as a project initialization and documentation generator. It analyzes existing project files to determine the best documentation approach (Agile, Waterfall, or Hybrid) and creates a standard structure.
Evidence:
Scans for standard project indicators like package.json, composer.json, and .git/ to identify tech stacks and development maturity.
Populates documentation templates in projects/<ProjectName>/_context/ based on user input and detected project characteristics.
Includes extensive industry-specific examples (healthcare, finance, etc.) that serve as minimal skeletons for documentation.
[COMMAND_EXECUTION]: The skill generates utility scripts for moving documentation files and references a document build script. These operations are transparent and perform localized tasks.
Evidence:
Generates export-docs.sh (bash) and export-docs.ps1 (PowerShell) to facilitate copying .docx deliverables into a project's export/ folder.
Skill instructions describe invoking scripts/build-doc.sh to assemble markdown files into final deliverables using local shell commands.
[PROMPT_INJECTION]: Analysis of the skill instructions and metadata reveals no attempts to override safety filters or manipulate the agent's core constraints.
Evidence:
Instructions are highly structured and focused on a professional SDLC workflow (PRIME methodology) without attempts to disregard safety protocols or extract system prompts.
[DATA_EXFILTRATION]: The skill scans the project environment for configuration files, but this information is used only to populate the local documentation context.
Evidence:
No network-related commands or external HTTP requests for data transmission to non-whitelisted domains were found in the skill logic or generated scripts.

meta-initialization