speckit-companion-living-adopt

Drafts and registers living specs for existing code areas in spec-kit projects.

1|Updated Nov 21, 2020
One-click install
npx skills add https://github.com/LuanDopke/persefone --skill speckit-companion-living-adopt-luandopke
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: speckit-companion-living-adopt
Source: https://github.com/LuanDopke/persefone/tree/main/.agents/skills/speckit-companion-living-adopt
Command: npx skills add https://github.com/LuanDopke/persefone --skill speckit-companion-living-adopt-luandopke

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Bringing existing brownfield code under specification-driven workflows normally requires hand-writing specs from scratch, which is slow and often skipped. This Skill reads a named code area's surface (exports, routes, props, signatures) and drafts honest, clearly-marked [DRAFT] living specs, then registers the capability so the rest of the Living Specs pipeline recognizes it. ## Core Features & Use Cases - Surface-first spec drafting: Reads only exported symbols, routes, and signatures to propose a small capability tree, writing requirement-and-scenario specs with confidence tags and [NEEDS CLARIFICATION] markers. - Flexible storage layouts: Supports central specs under capabilities/<name>/spec.md or colocated specs next to the code, with the layout chosen by the developer at a review gate. - Idempotent registry integration: Appends capabilities to living-specs.yml via a deterministic helper script and verifies resolver recognition, without touching unrelated specs. - Use Case: You inherit a billing module with no documentation. Point the Skill at src/billing/, confirm the proposed capability, and receive a draft living spec plus a registered capability ready for the fold-back pipeline. ## Quick Start Ask the assistant to adopt the src/billing/ directory into a living spec using a central layout.

Frequently Asked Questions about speckit-companion-living-adopt

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

FAQPage Schema
How do I create specs for existing legacy code?▼

Point the adoption command at a specific directory or file set, and it reads the code surface to propose capabilities and draft living specs. You confirm the proposal at a review gate before anything is written or registered.

What is the difference between central and colocated living specs?▼

Central specs live under capabilities/<name>/spec.md in one folder, while colocated specs sit next to the code they describe at <area root>/<name>.spec.md. The choice is the developer's and can be changed later.

Does this require a spec-kit project structure?▼

Yes, it requires a spec-kit project with a .specify/ directory containing the companion extension scripts such as register-capability.py and resolve-spec-paths.py. Missing helpers are reported and skipped rather than crashing.

Will it scan or rewrite my whole repository automatically?▼

No, adoption is opt-in and incremental. It only processes the areas you explicitly name, never scans the whole repo on its own, and re-running for an already-registered capability is a safe no-op.

Why are the generated specs marked as drafts?▼

Specs are drafted surface-first from exports and signatures rather than deep behavioral analysis, so they are marked [DRAFT] with low-confidence items tagged [inferred] or [NEEDS CLARIFICATION]. They are starting points for human review, not verified ground truth.