doc-brd-autopilot_skill

This skill automates BRD generation from reference docs or prompts, validates readiness, and manages indexing and parallel execution for streamlined planning.
  • 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-brd-autopilot

  • SKILL.md58.0 KB

Overview

This skill automates end-to-end Business Requirements Document (BRD) creation and review from reference documents (docs/00_REF/ or REF/) or user prompts. It detects input type, determines BRD category (platform vs feature), generates BRD content with real-time quality feedback, validates readiness, and maintains a BRD master index while supporting parallel execution. It orchestrates specialized sub-skills for naming, generation, validation, review, and automated fixes.

How this skill works

The autopilot parses the input to decide REVIEW mode (existing BRD) or GENERATE mode (reference files or prompt). It catalogs sources, validates @ref links, and runs a BRD-type questionnaire to choose Platform or Feature format. It reserves an ID, loads templates, generates document control and core sections (1–9) including mandatory ADR topics, and invokes a quality advisor during generation. After writing files it runs a validator, applies auto-fixes if needed, runs reviewer/fixer cycles and updates the BRD-00_index.

When to use it

  • Kickstarting a new project that needs a formal BRD from strategy or spec documents
  • Converting multiple technical or business reference docs into consistent BRDs
  • Automating BRD production in CI/CD or documentation pipelines
  • Creating both platform-level and feature-level BRDs with correct dependencies
  • Bulk-processing a folder of reference documents to produce many BRDs

Best practices

  • Keep authoritative source files in docs/00_REF/ or REF/ and validate @ref links before running generation
  • Provide clear stakeholder and project context in prompts to improve BRD quality
  • Run platform BRDs before dependent feature BRDs to preserve traceability
  • Use the reserved ID check to avoid collisions when running parallel jobs
  • Review PRD-Ready scores and iterate using the review/fixer cycle until score targets are met

Example use cases

  • Generate a BRD from a foundation technical specification in docs/00_REF/foundation/
  • Create a feature BRD by supplying a single prompt: --prompt "Create a BRD for user authentication system"
  • Review an existing BRD: provide BRD-01 to run the automated reviewer and get a quality report
  • Batch-generate BRDs for a module by running the autopilot on all files in docs/00_REF/
  • Integrate into CI to auto-validate and mark BRDs as PRD-ready when score thresholds are reached

FAQ

Reference files under docs/00_REF/ or REF/, the special argument all, or a user prompt via --prompt. Supplying BRD-NN runs review mode.

How does it decide Platform vs Feature BRD?

It runs an automated questionnaire and keyword analysis (title/content) and checks for platform dependencies; default is Platform when uncertain.

What happens if a referenced source is missing?

Mandatory missing references block generation with a clear error. Optional refs are logged as warnings and placeholders may be inserted.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
doc-brd-autopilot skill by vladm3105/aidoc-flow-framework | VeilStrat