OpenParable
Skip to content
← Back to skills

Specsafe New

Productivity
specsafe-new

Specsafe New is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Create a new spec with PRD, tech stack, and rules. Initializes SPEC stage.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/workflow/specsafe-new/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/specsafe-new, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Create a new SpecSafe specification (NEW → SPEC stage).

When to use:

  • Starting a new feature
  • Need to define requirements before coding
  • Want to establish tech stack and rules upfront

Input: Feature name or brief description

Steps

  1. Validate project state

    Check PROJECT_STATE.md exists (run specsafe init if not). Review current active specs to avoid conflicts.

  2. Create the spec

    specsafe new "<feature-name>"
    

    This:

    • Generates a new spec file in specs/active/
    • Assigns a unique SPEC-ID (e.g., SPEC-20260211-001)
    • Sets status to SPEC stage
    • Updates PROJECT_STATE.md
  3. Initialize PRD structure

    Open the new spec file and populate:

    • Purpose (WHY): One paragraph explaining why this feature exists
    • Scope (WHAT): In scope and out of scope sections
    • Tech Stack: Target frameworks, languages, tools
    • Rules: Coding standards, constraints, conventions
  4. Define initial requirements

    Add to Requirements table:

    • Functional requirements (what the system must do)
    • Non-functional requirements (performance, security)
    • Priority levels (P0=must have, P1=should have, P2=nice to have)
  5. Show next steps

    Display:

    • Spec file location
    • Current stage (SPEC)
    • Next command: /specsafe:spec <id> to flesh out details

Output

After creation:

  • ✅ Spec file created at specs/active/SPEC-YYYYMMDD-NNN-<feature-name>.md
  • ✅ Status set to SPEC stage
  • ✅ PROJECT_STATE.md updated
  • 📋 Prompt: "Edit the spec to add requirements, then run /specsafe:spec <id>"

Guardrails

  • Feature name should be descriptive (kebab-case recommended)
  • Check for naming conflicts with existing specs
  • Tech stack decisions should reference existing project standards
  • Never start coding without a spec in SPEC or later stage

Example

User: /specsafe:new "user-authentication"
→ Creates SPEC-20260211-004-user-authentication.md
→ Status: SPEC
→ Next: /specsafe:spec SPEC-20260211-004