- Home
- Skills
- Vladm3105
- Aidoc Flow Framework
- Doc Ears Autopilot
doc-ears-autopilot_skill
- Python
9
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill vladm3105/aidoc-flow-framework --skill doc-ears-autopilot- SKILL.md41.6 KB
Overview
This skill automates generation of formal EARS (Easy Approach to Requirements Syntax) statements from PRD documents and validates readiness for BDD workflows. It analyzes PRD content, categorizes requirements, produces structured EARS files in the required nested folder layout, and enforces BDD-Ready scoring and traceability. The autopilot delegates specialized checks and fixes to focused sub-skills while orchestrating the end-to-end pipeline.
How this skill works
The autopilot detects whether the input is a PRD or an existing EARS and chooses Generate or Review mode accordingly. For PRDs it runs a readiness validation, auto-fixes common gaps, extracts features and thresholds, categorizes requirements (Event, State, Unwanted, Ubiquitous), and generates EARS statements with IDs, traceability tags, and Quality Attributes sections. After generation it runs structural and BDD-Ready validation, applies fixes if needed, and produces nested EARS files and an index.
When to use it
- You have completed PRD documents and want automated EARS generation
- You need formal WHEN-THE-SHALL statements with timing and threshold traceability
- You require EARS and BDD-Ready score validation as part of delivery gates
- You want automatic categorization (Event/State/Unwanted/Ubiquitous) and quality attribute extraction
- You need output in the prescribed nested docs/03_EARS folder structure
Best practices
- Ensure PRDs achieve minimum EARS-Ready score (default 90%) before running generation
- Keep PRD sections (functional reqs, acceptance criteria, timing matrix) well-structured to improve extraction accuracy
- Use nested folder naming and element ID conventions (EARS-NN_{slug}) for consistent traceability
- Review quality-advisor feedback during generation to catch anti-patterns early
- Prefer the autopilot for bulk or standard EARS generation; use the direct EARS skill for heavy manual customization
Example use cases
- Generate EARS-05 from PRD-05 when EARS-05 is missing, auto-populating thresholds and QA tables
- Review and validate an existing EARS document (EARS-03) and produce a BDD-Ready score report
- Batch process multiple PRDs (PRD-01,PRD-02) to produce a set of standardized EARS files and a summary report
- Auto-fix missing timing matrices or state diagrams in PRDs before EARS generation
- Produce annotated traceability lists linking EARS statements to BRD and PRD source references
FAQ
The autopilot expects repository-style identifiers (PRD-NN or EARS-NN) and structured PRD documents in the docs/02_PRD folder; sectioned or monolithic PRDs are supported.
What triggers Generate vs Review mode?
If the input type is PRD-NN the autopilot looks for EARS-NN and generates it if missing; if the input is EARS-NN it runs review/validation only.
Can the autopilot auto-fix issues?
Yes. It applies configurable auto-fix templates for common PRD or EARS gaps (timing matrix, boundary matrices, state diagrams) and re-validates until thresholds are met.