spec-planning

Generate feature plans from NFRs and acceptance criteria into plan.md and eval_criteria.yaml.

20|Updated Feb 18, 2026
One-click install
npx skills add https://github.com/Accelerated-Innovation/governed-ai-delivery --skill spec-planning-accelerated-innovation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spec-planning
Source: https://github.com/Accelerated-Innovation/governed-ai-delivery/tree/main/agents/codex/skills/backend/spec-planning
Command: npx skills add https://github.com/Accelerated-Innovation/governed-ai-delivery --skill spec-planning-accelerated-innovation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Generates a concrete feature execution plan by translating NFRs and acceptance scenarios into actionable artifacts, ensuring alignment with architecture and evaluation criteria.

Core Features & Use Cases

  • Reads NFRs from features/<feature_name>/nfrs.md and acceptance criteria from features/<feature_name>/acceptance.feature to scope the feature.
  • Produces plan.md detailing tasks, components (ports, domain modules, adapters, API routes), test plans, LLM eval hooks, risks, and ADR status.
  • Generates an eval_criteria.yaml (and references to FIRST/Virtues and evaluation thresholds) to enable governed evaluation of the feature.

Quick Start

Use this skill by providing a feature name and its NFRs and acceptance criteria to generate plan.md and eval_criteria.yaml.

Frequently Asked Questions about spec-planning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a feature implementation plan from NFRs and acceptance criteria?

To generate a feature implementation plan from NFRs and acceptance criteria, provide a feature name along with its NFRs and acceptance scenarios. The skill translates these inputs into an actionable plan.md and eval_criteria.yaml.

What is included in a feature plan derived from architecture and evaluation standards?

A feature plan derived from architecture and evaluation standards includes detailed tasks, components like ports and adapters, test plans, LLM eval hooks, risks, and ADR status considerations to ensure governed implementation.

How do I create eval_criteria.yaml for software feature evaluation?

You create eval_criteria.yaml for software feature evaluation by deriving it from defined NFRs and acceptance criteria. The output includes evaluation thresholds and references to FIRST/Virtues standards to enable governed testing.

Do I need pre-existing NFR and acceptance files to plan feature implementations?

Yes, planning feature implementations requires reading inputs from features/<feature_name>/nfrs.md and features/<feature_name>/acceptance.feature. These files scope the feature before generating the execution plan.

What's the best way to align feature tasks with architecture and evaluation criteria?

The best way to align feature tasks with architecture and evaluation criteria is to translate NFRs and acceptance scenarios into a concrete plan. This ensures all tasks, components, and test plans adhere to architectural standards.

Why does my feature plan need ADR considerations and LLM eval hooks?

Feature plans need ADR considerations and LLM eval hooks to document architectural decisions and enable governed evaluation. Including these elements ensures the implementation addresses technical risks and meets evaluation thresholds.