spae-plan

Decomposes SPEC.md requirements into an atomic, acyclic task graph in PLAN.md.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/mystilleef/spae-framework --skill spae-plan-mystilleef
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: spae-plan
Source: https://github.com/mystilleef/spae-framework/tree/main/skills/spae-plan
Command: npx skills add https://github.com/mystilleef/spae-framework --skill spae-plan-mystilleef

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning a written specification into an executable plan is error-prone: tasks end up too large, requirements get dropped, and dependencies form cycles. This Skill converts a normalized SPEC.md into a verified, atomic task graph so downstream agents can execute work one verifiable task at a time. ## Core Features & Use Cases - Requirement-to-task decomposition: Maps every R-NNN requirement in SPEC.md to atomic tasks (T-001 onward) with dependencies, intent, acceptance criteria, and verification commands. - Graph validation loop: Verifies full requirement coverage, acyclic ordering, and forward-only dependencies before persisting, looping back to fix gaps automatically. - State persistence: Writes PLAN.md and initializes the task registry in STATE.json, advancing the workstream to the inspect phase for zero-knowledge resumption by any agent. - Use Case: After running the spec phase on a feature proposal, invoke this Skill to produce a dependency-ordered PLAN.md where each task is small enough to build and verify in a single cycle. ## Quick Start Run the plan phase to decompose the current SPEC.md into an atomic task plan in PLAN.md.

Frequently Asked Questions about spae-plan

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

FAQPage Schema
How do I break a software specification into atomic tasks?▼

Provide a normalized SPEC.md with R-NNN requirements, then run the plan phase. It decomposes the spec into T-NNN tasks with dependencies, acceptance criteria, and verification commands, validating full requirement coverage before writing PLAN.md.

How to generate a dependency-ordered task plan from requirements?▼

The plan phase reads SPEC.md, maps every requirement to task Satisfies fields, sorts tasks by dependency, risk, and vertical value, and verifies the graph has no cycles or forward-order violations before persisting PLAN.md.

What inputs does the SPAE plan phase require?▼

It requires .spae/current/SPEC.md to exist and STATE.json to report phase: plan. It reads repository source only for codebase fit and never reads VERIFY.md, which is reserved for later phases.

Can the plan phase modify source code or tests?▼

No. Its write scope is strictly limited to PLAN.md and STATE.json. Source code, tests, configuration, and SPEC.md are forbidden writes; repository code is inspected read-only for pattern fit.

What happens if a requirement is missing from the generated plan?▼

The verification loop detects any R-NNN gap, returns to the drafting step to add the missing task, and re-verifies. It exits only when every requirement maps to at least one task and the graph has no violations.