Agent skill

MoAI SPEC Workflow

by modu-ai in modu-ai/moai-adk

Manages SPEC documents for MoAI-ADK development, with GEARS or EARS requirement notation, acceptance criteria and a link into the Plan-Run-Sync workflow.

Apache-2.0Auto-check passedDevelopment

Install MoAI SPEC Workflow

skills CLI
$ npx skills add modu-ai/moai-adk --skill moai-workflow-spec -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install modu-ai/moai-adk moai-workflow-spec --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/modu-ai/moai-adk.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/moai-workflow-spec .claude/skills/moai-workflow-spec && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
moai-workflow-spec
GitHub stars
1.2k
Token cost
~5.1k tokens
SKILL.md length
2,133 words
Files
8 (incl. references)
Skills in repo
48
Repo updated
First seen
Licence
Apache-2.0

At a glance

Manages SPEC documents for MoAI-ADK development, with GEARS or EARS requirement notation, acceptance criteria and a link into the Plan-Run-Sync workflow.

  • Writing a SPEC document with structured requirements for a new feature
  • SKILL.md covers Quick Reference, Implementation Guide, Resources and SPEC Scope and Classification, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Defining acceptance criteria before implementation starts

What it does

The skill defines requirements in GEARS notation, the current form, with five patterns: ubiquitous, event-driven, state-driven, capability gate and event-detected, which replaces the removed IF/THEN form. Where, While and When modifiers can chain into one compound clause, and the subject can be any noun rather than only the system. EARS is kept as a legacy reference so older SPECs stay readable, and lint behavior follows the GEARS migration policy.

Each SPEC uses a three-file layout of spec.md, plan.md and acceptance.md. A four-step requirement clarification process includes assumption analysis, Git worktrees isolate SPECs developed in parallel, and TRUST 5 quality gates validate the result. Reference files cover an EARS deep dive, examples, migration, requirement clarification and the worktree workflow. The skill is designed for Claude Code and limits itself to file tools, Git, ls, wc, mkdir, grep and glob.

When your agent uses it

  • Writing a SPEC document with structured requirements for a new feature
  • Defining acceptance criteria before implementation starts
  • Migrating an older EARS-style SPEC to the GEARS notation
  • Running several SPECs in parallel in separate Git worktrees

Example prompts

  • “Create a SPEC for the invoice export feature with acceptance criteria.”
  • “Rewrite the requirements in SPEC-014 using the GEARS patterns.”
  • “Clarify the requirements for the notification service, listing my assumptions first.”
  • “Set up an isolated worktree for the payments SPEC so I can work on two SPECs at once.”

Requirements

  • MoAI-ADK, which the skill is written for
  • Git, for worktree-based SPEC isolation
  • Compatibility (from SKILL.md): Designed for Claude Code
  • Pre-approved tools (allowed-tools): Read, Write, Edit, Bash(git:*), Bash(ls:*), Bash(wc:*), Bash(mkdir:*), Grep, Glob

What it can do on your machine

Read from SKILL.md and the folder at commit a2a184a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Edit
    • Bash(git:*)
    • Bash(ls:*)
    • Bash(wc:*)
    • Bash(mkdir:*)
    • Grep
    • Glob

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • adk.mo.ai.kr

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

  • Compatibility

    Designed for Claude Code

    From compatibility in the SKILL.md frontmatter.

Context cost

MoAI SPEC Workflow loads about 5.1k tokens when it runs, and up to ~21k if it reads all its reference files. Until then it costs about 55 tokens; SKILL.md has 2,133 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~55
When it runs · the whole SKILL.md, loaded when a task matches
~5.1k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~21k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from modu-ai/moai-adk at commit a2a184a, republished under its Apache-2.0 licence (© modu-ai). 2,133 words, ~5,054 tokens.

Download SKILL.mdSave it as .claude/skills/moai-workflow-spec/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.
name
moai-workflow-spec
description
SPEC workflow orchestration with EARS format requirements, acceptance criteria, and Plan-Run-Sync integration for MoAI-ADK development. Use when creating SPEC documents or defining acceptance criteria.
allowed-tools
Read, Write, Edit, Bash(git:*), Bash(ls:*), Bash(wc:*), Bash(mkdir:*), Grep, Glob
compatibility
Designed for Claude Code
when_to_use
Use for SPEC workflow orchestration: EARS-format requirements, acceptance criteria, user stories, requirements gathering, planning, and Plan-Run-Sync…
license
Apache-2.0
user-invocable
false
metadata.version
1.2.0
metadata.category
workflow
metadata.status
active
metadata.updated
2026-01-08
metadata.modularized
true
metadata.tags
workflow, spec, ears, requirements, moai-adk, planning
metadata.author
MoAI-ADK Team

SPEC Workflow Management

Quick Reference

SPEC Workflow Orchestration using GEARS notation (current) — backed by the EARS legacy backward-compatibility window — for systematic requirement definition and Plan-Run-Sync workflow integration.

Lint behavior canonicalized per the GEARS migration policy.

Core Capabilities:

  • GEARS-Format Specifications (current): Five requirement patterns with the unified compound clause [Where ...][While ...][When ...] The <subject> shall <behavior> and a generalized <subject> (any noun, not only "the system")
  • EARS Legacy Reference: All EARS patterns preserved per the lint engine's backward-compatibility policy to keep pre-v3 SPECs (those authored before GEARS became canonical) readable
  • Requirement Clarification: Four-step systematic process with assumption analysis
  • SPEC Document Templates: Standardized 3-file structure (spec.md / plan.md / acceptance.md)
  • Plan-Run-Sync Integration: Seamless workflow connection
  • Parallel Development: Git Worktree-based SPEC isolation
  • Quality Gates: TRUST 5 framework validation

GEARS Five Patterns (current notation):

PatternGEARS form (current)EARS form (legacy)Notes
Ubiquitous"The <subject> shall <behavior>""The system shall <behavior>"<subject> may be any noun: system, component, service, agent, function, artifact
Event-driven"When <event-detected>, the <subject> shall <behavior>""WHEN <event>, the system shall <action>"Unchanged trigger semantics
State-driven"While <state>, the <subject> shall <behavior>""WHILE <state>, the system shall <action>"Unchanged — promoted as a first-class pattern
Capability gate"Where <capability / feature flag / static config>, the <subject> shall <behavior>""WHERE <feature exists>, the system shall <action>"Reframed — represents capability gate / feature flag / static config (no longer "Optional")
Event-detected (replaces IF/THEN)"When <undesired-condition-detected>, the <subject> shall <response>"IF <condition> THEN <action> [DEPRECATED — use WHEN <event-detected>]The IF/THEN modality was removed; describe the same intent as a detected event

Unified compound clause: **Where** <precondition> **While** <state> **When** <event> the <subject> shall <behavior> — any subset of the three modifiers may chain.

See GEARS notation reference.

IF/THEN deprecated callout: Authoring guidance previously used IF <condition> THEN <action> to describe state-conditioned behavior. In GEARS that intent is expressed as When <condition-detected> (event-detected form). The lint engine emits a LegacyEARSKeyword warning (non-strict) or error (moai spec lint --strict) on residual IF/THEN in new SPECs. The 6-month backward-compatibility window remains active for legacy SPECs.

Generalized subject substitution: GEARS replaces the hardcoded "the system" subject with <subject>, which may be any noun. Authors writing NEW SPECs MAY use the generalized form. Examples of valid non-"the system" subjects:

  • "The skill shall present GEARS as the primary notation." (Ubiquitous, <subject> = skill)
  • "The agent shall return a blocker report instead of prompting the user." (Ubiquitous, <subject> = agent)
  • "When a SPEC author opens the file, the component shall display the deprecation banner." (Event-driven, <subject> = component)

Pre-v3 SPECs (those authored before GEARS became canonical) keep "The system" as the default subject for readability; existing readers do not need to relearn the canonical phrase.

EARS Five Patterns (legacy — 6-month backward-compatibility window):

PatternFormatUse
Ubiquitous"The system shall always X"Always active
Event-Driven"WHEN event THEN action"Trigger-response
State-Driven"WHILE state, the system shall ..."Conditional behavior (use WHILE, not legacy IF/THEN)
Unwanted"The system shall not X"Prohibition
Optional"Where possible, provide X"Nice-to-have

The legacy IF/THEN modality is replaced by GEARS When <event-detected> — see callout above.

When to Use:

  • Feature planning and requirement definition
  • SPEC document creation and maintenance
  • Parallel feature development coordination
  • Quality assurance and validation planning
  • Requirements gathering from user story narratives

Quick Commands:

bash
/moai plan "user authentication system"                   # Create new SPEC
/moai plan "login" "signup"                              # Parallel SPECs
/moai plan "payment processing" --branch                  # New branch
/moai plan SPEC-001 "add OAuth support"                   # Update existing

Implementation Guide

Core Concepts

SPEC-First Development Philosophy:

  • EARS format ensures unambiguous requirements
  • Requirement clarification prevents scope creep
  • Systematic validation through test scenarios
  • Integration with DDD workflow for implementation
  • Quality gates enforce completion criteria
  • Constitution reference ensures project-wide consistency
Constitution Reference (SDD 2025 Standard)

Constitution defines the project DNA that all SPECs must respect. Before creating any SPEC, verify alignment with .moai/project/tech.md.

Constitution Components: Technology Stack, Naming Conventions, Forbidden Libraries, Architectural Patterns, Security Standards, Logging Standards.

Constitution Verification: All SPEC technology choices align with Constitution stack versions, no forbidden libraries, naming conventions respected, architectural boundaries preserved.

WHY: Constitution prevents architectural drift and ensures maintainability.

SPEC Workflow Stages
StageActivity
1User Input Analysis — parse natural-language feature description
2Requirement Clarification — 4-step systematic process
3EARS Pattern Application — structure requirements using five patterns
4Success Criteria Definition — establish completion metrics
5Test Scenario Generation — create verification test cases
6SPEC Document Generation — produce standardized markdown
GEARS Format (current)

GEARS (Generalized EARS) is the canonical SPEC notation as of v3.0.0. It preserves Ubiquitous / When (event-driven) / While (state-driven) and reframes Where as a capability gate. The legacy IF/THEN modality is replaced by When <event-detected>.

GEARS notation is exhaustively described in docs-site GEARS notation reference and the canonical GEARS migration policy record.

Compound clause example (with non-"the system" subject):

Where the project is initialized While strict mode is active When a SPEC author runs moai spec lint, the lint engine shall emit a LegacyEARSKeyword finding for every residual IF/THEN modality.

This example chains all three GEARS modifiers (Where, While, When) and uses <subject> = "lint engine" rather than "the system".

EARS Format (legacy — 6-month backward-compatibility window)

Five patterns cover all requirement types. Each pattern has a specific use case and test strategy. Pre-v3 SPECs (those authored before GEARS became canonical) continue to use EARS notation and remain valid per the lint engine's backward-compatibility policy.

See EARS deep dive with examples per pattern for use cases, examples, and test strategies for Ubiquitous, Event-Driven, State-Driven, Unwanted, and Optional requirements.

Requirement Clarification Process

5-step systematic process:

  • Step 0: Assumption Analysis (Philosopher Framework) — surface technical, business, team, integration assumptions
  • Step 0.5: Root Cause Analysis (Five Whys) — surface problem to root cause for problem-driven SPECs
  • Step 1: Scope Definition — supported methods, validation rules, failure handling, session management
  • Step 2: Constraint Extraction — performance, security, compatibility, scalability
  • Step 3: Success Criteria — coverage targets, response time percentiles, functional completion, quality gates
  • Step 4: Test Scenario Creation — normal, error, edge, security cases

See requirement clarification detailed workflow for assumption documentation templates and Five Whys application.

[NEEDS CLARIFICATION] Marker Convention

[NEEDS CLARIFICATION: <topic>] markers identify unresolved questions in plan.md and research.md that MUST be settled before Implementation Kickoff Approval (plan→run HUMAN GATE).

Placement: ONLY in plan.md and research.md (NEVER in spec.md or acceptance.md).

Format:

  • [NEEDS CLARIFICATION: <specific topic>] — inline marker for open questions
  • Each marker MUST be addressable via the orchestrator's question-channel capability before run-phase entry
  • plan-auditor detects unclarified markers and flags as "clarification gate" finding

3-Layer Distinction:

  • [NEEDS CLARIFICATION: <topic>] — plan/research artifact blocker (user Q required)
  • TODO — code-level implementation debt (no user Q needed)
  • @MX:TODO — code-level annotation for untested/incomplete code

Processing:

  • plan-auditor scans for [NEEDS CLARIFICATION] markers during audit
  • If any remain, plan-auditor recommends resolution before Implementation Kickoff Approval
  • Orchestrator runs question-channel rounds to resolve each marked topic (where the harness lacks it, carry the unresolved marker into the blocker report rather than letting the gate pass silently)
  • Implementation Kickoff Approval (mandatory human gate) proceeds only after all clarifications are resolved
Plan-Run-Sync Workflow Integration

PLAN (/moai plan): manager-spec analyzes input → EARS requirements → clarification → SPEC creation in .moai/specs/ → optional --branch.

RUN (/moai run): manager-develop loads SPEC → ANALYZE-PRESERVE-IMPROVE (DDD) or RED-GREEN-REFACTOR (TDD) per quality.yaml constitution.development_mode → moai-workflow-testing reference → per-spawn Agent(general-purpose) domain delegation → quality-gate validation (Stop hook / /moai gate).

SYNC (/moai sync): manager-docs synchronizes documentation → API docs from SPEC → README and architecture updates → CHANGELOG → version control commit.

Parallel Development with Git Worktree

Worktree provides isolated working directories per SPEC for parallel development without branch switching. Benefits: parallel development, clear ownership boundaries, dependency isolation, risk reduction.

See worktree workflow patterns for creation commands and team collaboration examples.


Resources

SPEC File Organization

The artifact set is not fixed — it is determined by the SPEC's Tier, classified in the plan phase before authoring begins. Read the tier off the canonical table (.claude/rules/moai/workflow/spec-workflow.md § SPEC Complexity Tier) rather than counting files here.

  • .moai/specs/SPEC-{ID}/spec.md — GEARS specification (EARS accepted during the legacy window). Every tier.
  • .moai/specs/SPEC-{ID}/plan.md — implementation plan, milestones, technical approach. Every tier.
  • .moai/specs/SPEC-{ID}/acceptance.md — acceptance criteria, Given-When-Then scenarios. Tier M and above; at Tier S the criteria live inline in spec.md §3.
  • .moai/specs/SPEC-{ID}/design.md and .moai/specs/SPEC-{ID}/research.md — system design and codebase research. Tier L only.

[HARD] A SPEC directory MUST contain every artifact its own tier names — and MUST NOT be judged incomplete for omitting one its tier does not name. A Tier S directory holding spec.md and plan.md alone is complete; adding acceptance.md there duplicates criteria that already live in spec.md §3, which is the drift this clause exists to prevent.

State files: .moai/state/last-session-state.json. Generated docs: .moai/docs/api-documentation.md.

Show full SKILL.md (785 more words)Show less
SPEC Metadata Schema

Canonical 12 required fields (enforced by the SPEC frontmatter lint rule): id, title, version, status, created, updated, author, priority, phase, module, lifecycle, tags.

Status enum (8 values): draft → in-progress → implemented → completed | superseded | archived | rejected. (planned is retained in the enum as legacy-optional — NOT in the active flow; no agent authors a draft → planned transition. See .claude/rules/moai/development/spec-frontmatter-schema.md § Status Enum.)

Optional fields: issue_number, depends_on, lint.skip, bc_id, tier (S/M/L LEAN tier).

Full schema at .claude/rules/moai/development/spec-frontmatter-schema.md (SSOT).

SPEC Lifecycle Management

Three lifecycle levels:

LevelDescriptionMaintenance
spec-firstSPEC discarded after implementationNone
spec-anchoredSPEC maintained alongside implementationQuarterly review
spec-as-sourceSPEC is single source of truth, only SPEC edited by humansChanges regenerate impl

Transitions: spec-first → spec-anchored when production-critical, spec-anchored → spec-as-source when compliance or regeneration workflow required. Downgrade requires explicit justification.

Quality Metrics

SPEC Quality Indicators: requirement clarity (all EARS patterns used), test coverage (all requirements have scenarios), constraint completeness, success criteria measurability.

Validation Checklist: All EARS requirements testable, no ambiguous language ("should", "might", "usually"), all error cases documented, performance targets quantified, security requirements OWASP-compliant.

Token Management
PhaseToken Budget
PLAN~30%
RUN~60%
SYNC~10%

Context Optimization: SPEC document persists in .moai/specs/. Session state in .moai/state/. Minimal context transfer through SPEC ID reference. Agent delegation reduces token overhead.


SPEC Scope and Classification

What Belongs in .moai/specs/

The .moai/specs/ directory is EXCLUSIVELY for SPEC documents that define features to be implemented.

Valid SPEC Content: feature requirements in EARS format, implementation plans with milestones, acceptance criteria with Given/When/Then scenarios, technical specifications for new functionality, user stories with clear deliverables.

SPEC Characteristics: forward-looking (what WILL be built), actionable, testable, structured (EARS).

What Does NOT Belong in .moai/specs/
Document TypeWhy Not SPECCorrect Location
Security AuditAnalyzes existing code.moai/reports/security-audit-{DATE}/
Performance ReportDocuments current metrics.moai/reports/performance-{DATE}/
Dependency AnalysisReviews existing dependencies.moai/reports/dependency-review-{DATE}/
Architecture OverviewDocuments current state.moai/docs/architecture.md
API ReferenceDocuments existing APIs.moai/docs/api-reference.md
Meeting NotesRecords decisions made.moai/reports/meeting-{DATE}/
RetrospectiveAnalyzes past work.moai/reports/retro-{DATE}/
Out of Scope Classification Rules

These routing rules decide what is out of scope for a SPEC document (and where it belongs instead). When authoring a SPEC's own exclusions section, express each excluded item as a ### Out of Scope — <topic> H3 sub-heading with - bullets so the section satisfies the OutOfScopeRule lint.

[HARD] Reports analyze what EXISTS → .moai/reports/. SPECs define what will be BUILT → .moai/specs/.

[HARD] Documentation explains HOW TO USE → .moai/docs/. SPECs define WHAT TO BUILD → .moai/specs/.


Works Well With

  • moai-foundation-core: SPEC-First DDD methodology and TRUST 5 framework
  • moai-workflow-testing: DDD implementation and test automation
  • moai-workflow-project: Project initialization and configuration
  • moai-workflow-worktree: Git Worktree management for parallel development
  • manager-spec: SPEC creation and requirement analysis agent
  • manager-develop: DDD/TDD implementation based on SPEC requirements
  • /moai gate skill (or sync-phase-quality-gate.sh Stop hook): TRUST 5 quality validation and gate enforcement (former manager-quality role)

For migration scenarios and validation scripts: references/migration-guide.md.


Version: 1.3.1 (skill body compression pass) Integration Status: Complete - Plan-Run-Sync workflow with SDD 2025 features

<!-- moai:evolvable-start id="rationalizations" -->

Common Rationalizations

RationalizationReality
"The SPEC is obvious, I can skip EARS format"EARS exists because obvious requirements are the first to be misinterpreted. The format forces disambiguation.
"Acceptance criteria are redundant with the requirements"Requirements describe intent. Acceptance criteria describe observable evidence. Both are needed.
"I will refine the SPEC during implementation"Late refinement means wasted implementation. SPEC is the cheap place to change your mind.
"Research is a nice-to-have, not a blocker"Skipping research produces SPECs that conflict with existing code. research.md prevents rework.
"Annotation cycle is just user friction"Annotation catches misunderstandings before code is written. It is the cheapest feedback loop in the pipeline.
"This SPEC is small, I do not need a separate file"Every SPEC is a persistent contract. In-message SPECs cannot be referenced by /moai run SPEC-XXX.
<!-- moai:evolvable-end -->
<!-- moai:evolvable-start id="red-flags" -->

Red Flags

  • Requirements written in imperative prose instead of EARS (WHEN X, SHALL Y)
  • Acceptance criteria phrased as subjective judgments ("feels fast", "looks clean")
  • SPEC document missing research.md sibling when modifying existing code
  • Annotation cycle skipped or reduced to a single-turn "looks good"
  • Requirements use "should" where they mean "shall" (optional vs mandatory ambiguity)
  • SPEC-ID not registered in .moai/specs/ directory
<!-- moai:evolvable-end -->
<!-- moai:evolvable-start id="verification" -->

Verification

  • SPEC file exists at .moai/specs/SPEC-XXX/spec.md with unique ID
  • Every requirement uses EARS keywords (WHEN, WHILE, WHERE, IF, SHALL)
  • Every acceptance criterion is observable (test output, file existence, metric threshold)
  • AC verification commands follow the plain-command form — see worktree-integration.md § Refused Commands in a Worktree-Isolated Session (measured guard boundary + authoring rule)
  • research.md exists when the SPEC touches existing code
  • Annotation cycle completed with explicit user approval marker
  • SPEC references existing SPEC-IDs it depends on or supersedes
  • Out of Scope section present to prevent scope creep — at least one ### Out of Scope — <topic> H3 sub-heading with a - bullet entry (satisfies the OutOfScopeRule lint)
<!-- moai:evolvable-end -->

© modu-ai, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 7 other files (references) in .claude/skills/moai-workflow-spec of modu-ai/moai-adk.

  • SKILL.md
  • modules/advanced-patterns.md
  • references/ears-deep-dive.md
  • references/examples.md
  • references/migration-guide.md
  • references/reference.md
  • references/requirement-clarification.md
  • references/worktree-workflow.md

Open the folder on GitHubat commit a2a184a

Compare with similar skills

MoAI SPEC Workflow next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

MoAI SPEC Workflow compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
MoAI SPEC Workflow this skillmodu-ai/moai-adk1.2k—~5.1kAutomated safety check: PassApache-2.0
Spec Writergarrytan/gstack136k—~14kAutomated safety check: NotesMIT
Voiceover-First DevelopmentDevin-AXIS/iPolloWork6.8k—~879Automated safety check: PassCustom licence
Interview-Driven Spec Writerposhan0126/dotclaude871—~804Automated safety check: PassMIT
Spec-Driven Developmentaddyosmani/agent-skills103k1 repos~3.2kAutomated safety check: PassMIT
CCPM Project Managementautomazeio/ccpm8.4k—~1.1kAutomated safety check: PassMIT

Similar skills

  • Spec Writer

    garrytan/gstack

    Converts a vague idea into a precise, executable spec in five phases, files it as an issue and can start an agent on it in a fresh worktree.

    136k GitHub stars~14k tokensUpdated today
    DevelopmentAuto-check: notes
  • Voiceover-First Development

    Devin-AXIS/iPolloWork

    Starts a feature as a demo narration instead of a PRD: you approve the script before any code, then the agent builds in a fresh worktree and opens a PR with proof.

    6.8k GitHub stars~879 tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Interview-Driven Spec Writer

    poshan0126/dotclaude

    Interviews you about scope, behavior, edge cases and verification, then writes a self-contained SPEC.md that a fresh session can implement without this conversation.

    871 GitHub stars~804 tokensUpdated 1 mo ago
    Agent WorkflowsAuto-check passed
  • Spec-Driven Development

    addyosmani/agent-skills

    Writes a structured specification before any code, moving through gated specify, plan, tasks and implement phases, with an optional capability map for multi-part requests.

    103k GitHub starsUsed in 1 repo~3.2k tokens
    DevelopmentAuto-check passed
  • Runs a spec-driven workflow from PRD to epic to GitHub issues to parallel agents, with status, standup and blocked-work reports from bundled scripts.

    8.4k GitHub stars~1.1k tokensUpdated 6 mo ago
    Product & Project ManagementAuto-check passed
  • Feature Spec Generator

    cashew-labs/libretto

    Researches the codebase and relevant docs, asks clarifying questions, then writes a spec sheet in specs/ for a significant feature or complex fix.

    904 GitHub stars~2.4k tokensUpdated 1 mo ago
    DevelopmentAuto-check passed

More from modu-ai/moai-adk

All 48 skills in this repo
  • Builds hand-editable SVG diagrams from computed layout coordinates, lints the source and renders a 2x PNG, with rules for when mermaid is the better choice.

    1.2k GitHub stars~5.2k tokensUpdated today
    Auto-check: notes
  • MoAI Foundation Core

    modu-ai/moai-adk

    Reference for MoAI-ADK's core development principles: TRUST 5 quality gates, SPEC-first domain-driven workflow, agent delegation and token budgeting.

    1.2k GitHub stars~5k tokensUpdated today
    Auto-check passed
  • MoAI TDD Workflow

    modu-ai/moai-adk

    Drives test-first development through the RED, GREEN, REFACTOR cycle, with a config switch that selects between TDD and a DDD workflow for existing code.

    1.2k GitHub stars~3.1k tokensUpdated today
    Auto-check passed
  • MoAI Worktree Management

    modu-ai/moai-adk

    Gives each SPEC its own Git worktree with a registry of active workspaces, base-branch sync and cleanup of merged ones, inside the MoAI-ADK workflow.

    1.2k GitHub stars~3.9k tokensUpdated today
    Auto-check passed
  • Watches a pull request's CI checks after creation, separates required from auxiliary failures, applies limited safe fixes and escalates anything semantic to you.

    1.2k GitHub stars~2.4k tokensUpdated today
    Auto-check: notes
  • Database guidance for PostgreSQL, MongoDB, Redis and Oracle plus Neon, Supabase and Firestore: schema design, indexing, query tuning and cloud database choice.

    1.2k GitHub stars~2.8k tokensUpdated today
    Auto-check passed

Works with

Questions about MoAI SPEC Workflow

What does MoAI SPEC Workflow do?

Manages SPEC documents for MoAI-ADK development, with GEARS or EARS requirement notation, acceptance criteria and a link into the Plan-Run-Sync workflow. The skill defines requirements in GEARS notation, the current form, with five patterns: ubiquitous, event-driven, state-driven, capability gate and event-detected, which replaces the removed IF/THEN form. Where, While and When modifiers can chain into one compound clause, and the subject can be any noun rather than only the system.

When should I use MoAI SPEC Workflow?

MoAI SPEC Workflow fits situations like: writing a SPEC document with structured requirements for a new feature; defining acceptance criteria before implementation starts; migrating an older EARS-style SPEC to the GEARS notation; running several SPECs in parallel in separate Git worktrees.

How do I install MoAI SPEC Workflow in Claude Code?

Run `npx skills add modu-ai/moai-adk --skill moai-workflow-spec -a claude-code`. Or copy the skill folder (.claude/skills/moai-workflow-spec in modu-ai/moai-adk) into .claude/skills/moai-workflow-spec in your project. Claude Code loads it when a task matches its description.

How do I install MoAI SPEC Workflow in Codex?

Run `npx skills add modu-ai/moai-adk --skill moai-workflow-spec -a codex`. Or copy the skill folder (.claude/skills/moai-workflow-spec in modu-ai/moai-adk) into .agents/skills/moai-workflow-spec in your project. Codex loads it when a task matches its description.

Can I use MoAI SPEC Workflow in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add modu-ai/moai-adk --skill moai-workflow-spec -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/moai-workflow-spec, .gemini/skills/moai-workflow-spec, .github/skills/moai-workflow-spec and .opencode/skills/moai-workflow-spec in your project.

What does MoAI SPEC Workflow need to run?

SKILL.md names no scripts, command-line tools or credentials: MoAI SPEC Workflow is instructions for the agent only. Our summary lists: MoAI-ADK, which the skill is written for; Git, for worktree-based SPEC isolation. Its frontmatter pre-approves these tools: Read, Write, Edit, Bash(git:*), Bash(ls:*), Bash(wc:*), Bash(mkdir:*), Grep, Glob. Compatibility (from SKILL.md): Designed for Claude Code.

Does MoAI SPEC Workflow access the network?

SKILL.md names 1 domain. As links in the text: adk.mo.ai.kr. This is read from the text; nothing was executed.

Is MoAI SPEC Workflow safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does MoAI SPEC Workflow use?

MoAI SPEC Workflow is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does MoAI SPEC Workflow use?

About 5.1k tokens (SKILL.md is roughly 20k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 16k tokens, read only when the agent opens those files.

What are the alternatives to MoAI SPEC Workflow?

Skills that share tags, products or a category with MoAI SPEC Workflow: Spec Writer (garrytan/gstack, 136k stars), Voiceover-First Development (Devin-AXIS/iPolloWork, 6.8k stars), Interview-Driven Spec Writer (poshan0126/dotclaude, 871 stars) and Spec-Driven Development (addyosmani/agent-skills, 103k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains MoAI SPEC Workflow?

modu-ai (a GitHub organization) maintains it in modu-ai/moai-adk, which has 1,230 GitHub stars. The repository holds 48 skills in this directory. The repository was last updated on October 8, 2026.

Source: modu-ai/moai-adk on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.