tlc-spec-driven

Plan and implement software projects with traceable specifications and adaptive execution phases.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/luizgmassa/Useful-Agent-Skills --skill tlc-spec-driven-luizgmassa
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tlc-spec-driven
Source: https://github.com/luizgmassa/Useful-Agent-Skills/tree/main/skills/tlc-spec-driven
Command: npx skills add https://github.com/luizgmassa/Useful-Agent-Skills --skill tlc-spec-driven-luizgmassa

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps AI agents plan and implement software changes with clear requirements, correct depth based on complexity, and verification-first execution so work doesn’t drift, become untestable, or lose context between sessions.

Core Features & Use Cases

  • 4 adaptive phases (Specify → Design → Tasks → Execute) with auto-skipping to match scope size and complexity, from quick bug fixes to complex, ambiguous domains.
  • Requirement traceability and atomic task execution using requirement IDs, explicit “done when” criteria, and atomic git commits per task.
  • Persistent session memory and resumption via project state and pause/resume handoff docs, so decisions and progress survive across sessions.
  • Stack-agnostic workflow that supports any coding agent and delegates research/diagrams/code exploration when complementary skills are available.

Quick Start

Say: Initialize project, then specify feature user-authentication, execute the tasks, and pause work if anything needs a follow-up decision.

Frequently Asked Questions about tlc-spec-driven

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

FAQPage Schema
How do I maintain requirement traceability across complex software changes?

To maintain requirement traceability, map each feature specification to explicit requirement IDs and enforce atomic git commits per task. This creates a verifiable chain linking code changes directly to originating requirements.

What is the best way to map an existing codebase before planning new features?

Mapping an existing codebase involves analyzing current structures to create a baseline specification. This workflow supports codebase exploration to inform feature design and break down implementation into surgical file changes.

How do I resume a paused software implementation session without losing context?

To resume a paused software implementation session without losing context, generate and persist session handoff docs. This preserves project state and execution decisions, allowing agents to pick up exactly where they left off.

Can I use AI agents for software planning without locking into a specific tech stack?

Yes, you can use AI agents for software planning without a specific tech stack. This workflow is stack-agnostic, adapting execution phases to any coding agent and delegating technical research when needed.

How does adaptive execution handle small bug fixes versus complex domain changes?

Adaptive execution handles small bug fixes and complex domain changes by auto-skipping unnecessary workflow phases. It scales from quick surgical file changes to full specify, design, tasks, and execute phases.

What limitations exist when relying on AI agents for implementation verification?

When relying on AI agents for implementation verification, limitations include potential context drift without guardrails. This workflow mitigates it using a knowledge verification chain and optional interactive UAT to validate results.