How do I install the Using Spdd skill?

Run npx skillmds@latest add danmestas/using-spdd in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Using Spdd safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Using Spdd?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Using Spdd free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Using Spdd?

danmestas (@danmestas) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Using Spdd by danmestas — Agent Skill

Name: Using Spdd
Author: danmestas

File contents

Using Structured-Prompt-Driven Development

Structured-Prompt-Driven Development (SPDD) treats prompts as first-class delivery artifacts: versioned, reviewed, reused, and kept in sync with code.

Announce at start: "I'm using the using-spdd skill to keep prompt assets and code in a closed loop."

Fit check

Use SPDD for logic-heavy, audit-heavy, team-shared, or repeatable delivery work where intent drift is expensive. Avoid it for emergency recovery, throwaway spikes, one-off scripts, or taste-led creative work unless the user explicitly wants the governance overhead.

Non-negotiable loop

Clarify intent with spdd-alignment before design or code.
Analyze context with spdd-analysis against only relevant code and domain artifacts.
Write a REASONS Canvas with spdd-reasons-canvas before implementation.
Generate/modify code from the canvas with spdd-generate; no scope beyond Operations, Norms, and Safeguards.
Validate behavior first, then review structure with spdd-iterative-review.
If behavior or business logic changes: update the prompt first with spdd-prompt-update, then update code.
If only internal structure changes: refactor code, prove behavior is unchanged, then sync back with spdd-sync.

Artifact discipline

Default location: docs/superpowers/spdd/YYYY-MM-DD-<slug>/.

A complete SPDD slice has:

story.md or linked requirement source.
analysis.md for domain concepts, risks, scope, and strategy.
reasons-canvas.md as the executable prompt/spec.
verification.md or test script notes mapping behavior to acceptance criteria.

Keep artifacts current in the same branch as the code they govern. Never let chat-only instructions replace the canvas.

Prompt/code divergence rule

When reality diverges, classify the divergence before changing anything:

Logic correction / requirement change / acceptance mismatch: prompt first, code second.
Refactor / naming / decomposition with no observable behavior change: code first, prompt sync second.
Unclear intent: stop implementation and return to alignment.

Required review stance

Reviewers should check the intent artifact before the diff. A code change is not ready if the code passes tests but the canvas no longer describes what the system does.

---name: using-spdddescription: Use when a session is working under Structured-Prompt-Driven Development (SPDD), when prompts/specs must be versioned delivery artifacts, or when logic-heavy changes need a prompt-first closed loop.---# Using Structured-Prompt-Driven DevelopmentStructured-Prompt-Driven Development (SPDD) treats prompts as first-class delivery artifacts: versioned, reviewed, reused, and kept in sync with code.**Announce at start:** "I'm using the using-spdd skill to keep prompt assets and code in a closed loop."## Fit checkUse SPDD for logic-heavy, audit-heavy, team-shared, or repeatable delivery work where intent drift is expensive. Avoid it for emergency recovery, throwaway spikes, one-off scripts, or taste-led creative work unless the user explicitly wants the governance overhead.## Non-negotiable loop1. **Clarify intent** with `spdd-alignment` before design or code.2. **Analyze context** with `spdd-analysis` against only relevant code and domain artifacts.3. **Write a REASONS Canvas** with `spdd-reasons-canvas` before implementation.4. **Generate/modify code from the canvas** with `spdd-generate`; no scope beyond Operations, Norms, and Safeguards.5. **Validate behavior first**, then review structure with `spdd-iterative-review`.6. **If behavior or business logic changes:** update the prompt first with `spdd-prompt-update`, then update code.7. **If only internal structure changes:** refactor code, prove behavior is unchanged, then sync back with `spdd-sync`.## Artifact disciplineDefault location: `docs/superpowers/spdd/YYYY-MM-DD-<slug>/`.A complete SPDD slice has:- `story.md` or linked requirement source.- `analysis.md` for domain concepts, risks, scope, and strategy.- `reasons-canvas.md` as the executable prompt/spec.- `verification.md` or test script notes mapping behavior to acceptance criteria.Keep artifacts current in the same branch as the code they govern. Never let chat-only instructions replace the canvas.## Prompt/code divergence ruleWhen reality diverges, classify the divergence before changing anything:- **Logic correction / requirement change / acceptance mismatch:** prompt first, code second.- **Refactor / naming / decomposition with no observable behavior change:** code first, prompt sync second.- **Unclear intent:** stop implementation and return to alignment.## Required review stanceReviewers should check the intent artifact before the diff. A code change is not ready if the code passes tests but the canvas no longer describes what the system does.

Frequently asked questions

Using Spdd