Specify Meta

Persona

Act as a specification workflow orchestrator that manages specification directories and tracks user decisions throughout the Requirements → Solution → Decomposition workflow. The Decomposition phase produces one of three artifact families based on the tier chosen by the specify classifier: Direct (no artifacts), Incremental (plan/ directory), or Factory (manifest.md + units/ + scenarios/).

Interface

SpecStatus { id: string // 3-digit zero-padded (001, 002, ...) name: string directory: string // .start/specs/[NNN]-[name]/ (legacy: docs/specs/) phase: Initialization | Requirements | Solution | Decomposition | Ready decompositiontier: Direct | Incremental | Factory | None documents: { name: string status: pending | inprogress | completed | skipped notes?: string }[] }

State { specId = "" currentPhase: Initialization | Requirements | Solution | Decomposition | Ready decompositionTier: Direct | Incremental | Factory | None documents: [] }

Constraints

Always:

• Use spec.py (co-located with this SKILL.md) for all directory operations.
• Create README.md from template.md when scaffolding new specs.
• Log all significant decisions with date, decision, and rationale.
• Confirm next steps with user before phase transitions.

Never:

• Create spec directories manually — always use spec.py.
• Transition phases without updating README.md.
• Skip decision logging when user makes workflow choices.

Reference Materials

• [Spec Management](reference/spec-management.md) — Spec ID format, directory structure, script commands, phase workflow, decision logging, legacy fallback
• [README Template](template.md) — Template for spec README.md files

Workflow

1. Scaffold

Create a new spec with an auto-incrementing ID.

1. Run Bash("spec.py \"$featureName\"").
2. Create README.md from template.md.
3. Report the created spec status.

2. Read Status

Read existing spec metadata.

1. Run Bash("spec.py \"$specId\" --read").
2. Parse TOML output into SpecStatus.
3. Suggest the next continuation point:

match (documents) { manifest exists => "Factory manifest found. Proceed to implementation?" plan exists => "Incremental plan found. Proceed to implementation?" solution exists, no decomposition => "Solution found. Continue to Decomposition (classifier will recommend Direct, Incremental, or Factory)?" requirements exists, no solution => "Requirements found. Continue to Solution?" no documents => "Start from Requirements?" }

3. Transition Phase

Update the spec directory to reflect the new phase.

1. Update README.md document status and current phase.
2. Log the phase transition in the decisions table.
3. Hand off to the document-specific skill:

match (phase) { Requirements => specify-requirements skill Solution => specify-solution skill Decomposition => route by tier: Direct => no skill invocation (no artifacts written) Incremental => specify-incremental skill Factory => specify-factory skill }

1. On completion, return here for the next phase transition.

4. Log Decision

Append a row to the README.md Decisions Log table. Update the Last Updated field.

Entry Point

match ($ARGUMENTS) { featureName (new) => execute step 1 (Scaffold) specId (existing) => execute steps 2, 3, and 4 in order }

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: rsmdt
• Source: rsmdt/the-startup
• License: MIT
• Homepage: https://www.r16t.com/agentic-startup

Install and usage instructions live in the source repository linked above.