- Home
- Skills
- Prompt & Skill Engineering
- spec/plan driven development
spec/plan driven development
Maintain strict consistency between requirements, plans, tasks, and verification using a four-document SDD workflow.
$1.99
Works with the AI Agents you already use
spec/plan driven development
Example session with this skill installed
Initialize the spec-driven development documents for a new JWT authentication module in the current repository.
- Read your context and instructions
- Compiled the spec plan driven
I have established the project facts and generated the Four-Document Set for the JWT authentication module in spec/jwt-auth/. These documents define the requirements, technical strategy, and implementation path.
Specification and Planning Documents
| File | Responsibility | Status |
|---|---|---|
spec/jwt-auth/spec.md | Requirements F*, N* and Acceptance Criteria AC* | Drafted |
spec/jwt-auth/plan.md | Implementation strategy and data flow | Drafted |
spec/jwt-auth/task.md | Execution steps T* with dependency tracking | Initialized |
spec/jwt-auth/checklist.md | Verification items C* mapped to AC* | Pending |
Initial Project Context
- Target Directory:
spec/jwt-auth/ - Detected Stack: Node.js, Express, Vitest.
- Scope: Token generation, middleware validation, and secret rotation.
Next steps
- Review the
spec.mdto ensure the functional requirements match your expectations. - Approve the technical strategy in
plan.mdbefore I begin the implementation tasks. - Run
ls spec/jwt-auth/to verify all four documents are present in your local workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Software development often suffers from drift between requirements, design, and implementation. This skill enforces a structured Specification-Driven Development (SDD) workflow by generating and synchronizing a "Four-Document Set": spec.md for requirements, plan.md for technical strategy, task.md for execution, and checklist.md for verification. It prevents scope creep and ensures every line of code traces back to a validated user need.
What it does
- Requirement Analysis: Generates
spec.mddefining goals, functional/non-functional requirements, and clear acceptance criteria. - Technical Planning: Drafts
plan.mdoutlining data flow, module boundaries, and architectural decisions without bloating it with implementation code. - Task Decomposition: Breaks work into atomic
task.mdentries linked to specific requirements and dependencies. - Verification Mapping: Creates a
checklist.mdthat maps manual and automated tests directly to acceptance criteria. - Progress Recovery: Reconstructs development state from existing documents and code when resuming interrupted sessions.
How it works
- Fact-Finding: The agent scans the target directory to identify existing frameworks, languages, and build tools.
- Drafting: It generates the four-document set in a dedicated
spec/<feature-name>/directory. - Execution: You authorize specific tasks; the agent updates task statuses, records evidence, and performs validations.
- Consistency Audit: It checks for circular dependencies, missing acceptance tests, or requirements without implementation tasks.
Frameworks & tools
This skill is framework-agnostic. It adapts to existing project structures, build systems, and testing suites by observing the repository's native patterns and configuration files.
Why this beats prompting it yourself
Manual prompting often loses track of the "why" behind code changes over long sessions. This skill mandates a bidirectional link between requirements and verification, ensuring that documentation is an accurate reflection of the codebase rather than an outdated artifact.
Use cases
- Drafting a formal implementation plan before writing any code.
- Synchronizing technical tasks after a major change in project requirements.
- Auditing an existing feature to ensure all acceptance criteria are met.
- Resuming complex development workflows after a session timeout or context switch.
Known limitations
Does not automatically generate four-document sets for simple code explanations or minor edits unless explicitly requested. Requires an existing or planned directory structure for documentation storage.
Will not trigger for simple one-off code edits unless requested. Requires a structured directory like spec/feature-name/ to maintain document sets.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 25 days ago
Permissions
No special permissions declared or detected