What problem does it solve?
This Skill simplifies the complex and error-prone task of extracting specific data points from diverse .runs/ artifacts (Markdown, JSON, YAML) for auditing, reporting, or automated decision-making. It replaces manual grep/wc pipelines with deterministic, null-safe operations.
Core Features & Use Cases
- Deterministic Extraction: Provides reliable counting (e.g., requirements, BDD scenarios, open questions) and field extraction from JSON receipts, YAML blocks, and Machine Summary sections.
- Read-Only Operations: Ensures data integrity and auditability by strictly reading artifacts, never modifying them.
- Unified & Simplified Interface: Offers a consistent, shimmed command-line interface for various parsing tasks, abstracting away underlying complexity.
- Use Case: After a "Signal" flow, use this skill to quickly count the number of functional requirements (REQ- markers) and open questions (QID- markers) from the generated
requirements.md and open_questions.md files, providing a precise summary for review and subsequent flows.
Quick Start
Use the runs-derive skill to count all lines starting with '### REQ-' in the file '.runs/my-feature/signal/requirements.md', returning the count.
