mcpField
W
workflow-optimizer

workflow-optimizer

github.com/vladm3105Agent workflows1754 installsUpdated 1d ago
Repository

About workflow-optimizer

workflow-optimizer is a skill for teams using an eight-layer SDD workflow. It maps project artifacts to workflow status, recommends prioritized next steps, identifies parallel work, and reports progress.

Key features

  1. 01Artifact status and layer tracking
  2. 02Prioritized next-step recommendations
  3. 03Parallel-work identification
  4. 04Progress and critical-path reporting

Use cases

  1. 01Choosing what artifact to create next
  2. 02Reviewing project workflow progress
  3. 03Finding work that can proceed in parallel

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/vladm3105/aidoc-flow-framework --skill workflow-optimizer

After installing, type/workflow-optimizerin Claude Code to confirm it loaded.

Details

From the project README

workflow-optimizer

Purpose

Guide a user through the 8-layer SDD workflow (BRD → PRD → EARS → BDD → ADR → SPEC → TDD → IPLAN → Code). It locates the current position from completed artifacts, recommends prioritized next steps with rationale, surfaces work that can proceed in parallel, and reports progress — removing the friction of manually deciding what to build next.

When to Use

Use workflow-optimizer when:

  • You just completed an artifact and need next-step guidance.
  • You are starting documentation and want a workflow overview.
  • You want to find parallel-work opportunities or a progress report.

Do not use it to pick a skill for a specific task (use ../skill-recommender/SKILL.md), to gather project context (use ../context-analyzer/SKILL.md), or to validate artifacts (use ../trace-check/SKILL.md or ../quality-advisor/SKILL.md).

Behavior

Given a project_root (and optionally a just-completed artifact and a focus filter), the skill:

  1. Analyzes project state — discovers artifacts and extracts status (Draft, In Review, Approved, Superseded, Deprecated) from Document Control, producing counts by type.
  2. Determines workflow position — maps artifacts to layers 1–8 (each with a single-layer prerequisite per the cumulative chain), marking layers as completed, in-progress, ready, or blocked, and computes a progress percentage.
  3. Identifies required next steps — uses the dependency chain to list mandatory next artifacts, prioritized P0/P1/P2 with rationale and what each unblocks.
  4. Finds parallel opportunities — independent BRD/PRD tracks, ADRs that can be drafted concurrently, and BDD scenarios that can start once EARS requirements exist; notes items blocked by missing upstream layers.
  5. Calculates progress metrics — per-layer status and the critical path (EARS → BDD → ADR → SPEC → TDD → IPLAN).
  6. Generates recommendations — concrete next actions naming the layer skill to run (e.g. ../doc-ears/SKILL.md), parallel suggestions, blocked items with unblock paths, and a progress summary.

Related Resources

  • Layer registry (layers & prerequisites): framework/registry/LAYER_REGISTRY.yaml
  • Traceability: framework/governance/TRACEABILITY.md
  • Layer skills: ../doc-brd/SKILL.md … ../doc-iplan/SKILL.md
  • Related skills: ../context-analyzer/SKILL.md · ../skill-recommender/SKILL.md · ../doc-flow/SKILL.md

Comments

Markdown supported