Skip to content
/

cline-spec-kit-workflows (JRedeker)

cline-spec-kit-workflows · JRedeker/cline-spec-kit-workflows · ★ 16 · last commit 2025-10-12

Ports GitHub Spec Kit's spec-driven development workflow to Cline via eight procedural workflow files.

Best whenImplementation should be blocked until all pre-phase checklists pass; checklist-count logic (counting - [X] vs - [ ] patterns) should be the gate condition.
Skip ifWriting code before spec is complete, Running create-new-feature.sh more than once per feature
vs seeds
spec-kitseed (same artifact sequence, same GitHub Spec Kit origin), but targets Cline instead of GitHub Copilot, and acts as an …
Primitive shape 8 total
Commands 8
00

Summary

cline-spec-kit-workflows — Summary

cline-spec-kit-workflows ports GitHub's Spec Kit spec-driven development workflow to Cline via eight .clinerules/workflows/ markdown files. The workflows cover a complete SDD lifecycle: Constitution → Specify → Clarify → Plan → Checklist → Tasks → Analyze → Implement. Each workflow file is a detailed procedural prompt that calls Bash scripts (from the .specify/scripts/ system), manages YAML front matter for Cline command integration, and produces structured artifacts in a .specify/ directory tree. The "specify" workflow alone runs quality-validation checklists, enforces a maximum of 3 [NEEDS CLARIFICATION] markers, and writes checklist files that must all pass before implementation proceeds. This is a pure documentation-layer framework — no code beyond the workflow prompt files is shipped in this repository.

differs_from_seeds: Most similar to spec-kit (Archetype 2 — mirror commands + skills), sharing the exact same artifact structure (constitution, specification, plan, tasks, implement) and even referencing the GitHub Spec Kit project explicitly. However, cline-spec-kit-workflows differs from the spec-kit seed in two ways: (1) it targets Cline specifically (.clinerules directory format) rather than GitHub Copilot, and (2) it relies on external Bash scripts from the .specify/ system (which is not shipped in this repo — users must install GitHub Spec Kit separately) rather than shipping its own scripts. Compared to openspec, which has its own script runner, this framework is a Cline-specific adaptation layer over an existing spec system.

01

Overview

cline-spec-kit-workflows — Overview

Origin

Created by JRedeker as "Cline workflow files for GitHub Spec Kit - comprehensive workflows for spec-driven development!" The project bridges two existing tools: GitHub's Spec Kit (a spec-driven development methodology) and Cline (the popular VS Code AI coding assistant).

Philosophy

The README states the core workflow philosophy:

"These workflows enable structured, methodical software development by guiding you through a complete specification-driven process"

Explicitly listed benefits:

  • Reduced Rework: Catch issues early in the specification phase
  • Clear Communication: Shared understanding between humans and AI
  • Quality Assurance: Built-in validation and consistency checking
  • Structured Progress: Clear milestones from idea to implementation
  • Knowledge Capture: Documentation naturally emerges from the process
  • Risk Mitigation: Identify and address uncertainties before coding

Workflow Ordering Philosophy

"Note: These workflows are designed to enhance your development process, not replace human judgment. Use them as guides and adapt them to your specific needs and context."

The constitution phase is explicitly a "one-time setup" — establishing project principles before any feature work. This mirrors the concept of a project charter in traditional project management.

Quality Gate Philosophy (from specify.md)

The specify workflow limits AI speculation:

"Limit: Maximum 3 [NEEDS CLARIFICATION] markers total. Prioritize clarifications by impact: scope > security/privacy > user experience > technical details"

And enforces validation:

"If items fail (excluding [NEEDS CLARIFICATION]): [auto-repair loop]"

Integration with GitHub Spec Kit

This repo is explicitly positioned as an integration layer:

"These workflows are designed to work seamlessly with GitHub Spec Kit. When used together, they provide: Consistent artifact structure, Version-controlled specifications, Clear audit trail of decisions, Easy onboarding for new team members, Integrated planning and execution"

02

Architecture

cline-spec-kit-workflows — Architecture

Distribution

Standalone repository. Install by copying .clinerules/ directory to project root.

Install

cp -r .clinerules /path/to/your/project/

Optional: adapt for other AI assistants (workflows are plain markdown).

Directory Tree

cline-spec-kit-workflows/
├── .clinerules/
│   └── workflows/
│       ├── constitution.md    # One-time project principles setup
│       ├── specify.md         # Feature specification creation
│       ├── clarify.md         # Structured ambiguity resolution
│       ├── plan.md            # Implementation plan generation
│       ├── checklist.md       # Quality validation checklist
│       ├── tasks.md           # Task breakdown
│       ├── analyze.md         # Cross-artifact consistency check
│       └── implement.md       # Implementation execution
├── .gitignore
├── LICENSE
└── README.md

Rule File Format

Cline .clinerules format with YAML front matter:

---
description: Create or update the feature specification from a natural language feature description.
---

The workflow files use:

  • YAML front matter with description field (used by Cline as command metadata)
  • $ARGUMENTS placeholder (Cline substitutes the text after the command invocation)
  • References to external Bash scripts via Run script .specify/scripts/bash/<script>.sh
  • Inline artifact templates (spec quality checklists, etc.)

External Dependencies

Critical: The workflow files reference Bash scripts from the GitHub Spec Kit system (.specify/scripts/bash/). These scripts are NOT shipped in this repository — they come from the GitHub Spec Kit installation. This means:

  1. Users must install GitHub Spec Kit separately
  2. The .specify/ directory structure must exist in the project
  3. The workflow files are useless standalone without the Spec Kit backend

Target AI Tools

Primary: Cline (VS Code extension) Secondary: Any AI assistant that supports markdown workflows

Required Runtime

  • Cline VS Code extension
  • GitHub Spec Kit (for the .specify/scripts/ Bash scripts)
  • Bash (for script execution)
03

Components

cline-spec-kit-workflows — Components

Workflow Files (8 total)

Name File Phase Optional? Purpose
constitution constitution.md Setup No (one-time) Establish project principles, constraints, governance; populate .specify/memory/constitution.md; propagate to all templates
specify specify.md Specify No Create feature spec from natural language description; run quality validation; write to SPEC_FILE
clarify clarify.md Clarify Yes Structured questioning to resolve ambiguities; produce resolved-questions document
plan plan.md Plan No Generate implementation plan from spec; produce plan.md artifact
checklist checklist.md Validate Yes Generate quality validation checklists for requirements
tasks tasks.md Tasks No Break plan into concrete, actionable tasks; produce tasks.md with completion criteria
analyze analyze.md Validate Yes Cross-artifact consistency check; identify contradictions between spec/plan/tasks
implement implement.md Implement No Execute tasks from tasks.md; verify checklist pass/fail before proceeding; manage ignore files

External Script Dependencies (from GitHub Spec Kit)

Script Called By Purpose
.specify/scripts/bash/create-new-feature.sh specify.md Creates feature directory, returns BRANCH_NAME and SPEC_FILE as JSON
.specify/scripts/bash/check-prerequisites.sh implement.md Validates prerequisites, returns FEATURE_DIR and AVAILABLE_DOCS

Artifact Files Written

Artifact Written By Path
constitution constitution.md .specify/memory/constitution.md
spec specify.md SPEC_FILE (from create-new-feature.sh output)
spec quality checklist specify.md FEATURE_DIR/checklists/requirements.md
plan plan.md FEATURE_DIR/plan.md
tasks tasks.md FEATURE_DIR/tasks.md

Templates (Inline in Workflow Files)

The workflow files contain inline templates for:

  • Spec quality checklist (inside specify.md)
  • Sync Impact Report format (inside constitution.md)
  • Various checklist structures (inside checklist.md)
05

Prompts

cline-spec-kit-workflows — Prompts

Prompt 1: specify.md (Feature Specification)

Technique: Scripted procedural prompt with numbered execution flow, inline template injection, and quality gate loop.

---
description: Create or update the feature specification from a natural language feature description.
---

## User Input

```text
$ARGUMENTS

You MUST consider the user input before proceeding (if not empty).

Outline

...Given that feature description, do this:

  1. Run the script .specify/scripts/bash/create-new-feature.sh --json "$ARGUMENTS" from repo root and parse its JSON output for BRANCH_NAME and SPEC_FILE. IMPORTANT You must only ever run this script once.

  2. Follow this execution flow:

    1. Parse user description from Input If empty: ERROR "No feature description provided"
    2. Extract key concepts from description Identify: actors, actions, data, constraints
    3. For unclear aspects:
      • Make informed guesses based on context and industry standards
      • Only mark with [NEEDS CLARIFICATION: specific question] if:
        • The choice significantly impacts feature scope or user experience
        • Multiple reasonable interpretations exist with different implications
        • No reasonable default exists
      • LIMIT: Maximum 3 [NEEDS CLARIFICATION] markers total
      • Prioritize clarifications by impact: scope > security/privacy > user experience

**Technique**: Procedural scripted execution with error handling (ERROR keyword), strict limits on clarification requests (max 3), and inline checklist template injection. The `$ARGUMENTS` placeholder is Cline's standard argument substitution mechanism.

---

## Prompt 2: constitution.md (Project Constitution)

**Technique**: Template management + cascading propagation + semantic versioning enforcement.

```markdown
You are updating the project constitution at `.specify/memory/constitution.md`. 
This file is a TEMPLATE containing placeholder tokens in square brackets 
(e.g. `[PROJECT_NAME]`, `[PRINCIPLE_1_NAME]`). Your job is to (a) collect/derive 
concrete values, (b) fill the template precisely, and (c) propagate any amendments 
across dependent artifacts.

`CONSTITUTION_VERSION` must increment according to semantic versioning rules:
* MAJOR: Backward incompatible governance/principle removals or redefinitions.
* MINOR: New principle/section added or materially expanded guidance.
* PATCH: Clarifications, wording, typo fixes, non-semantic refinements.

6. Validation before final output:
   - No remaining unexplained bracket tokens.
   - Version line matches report.
   - Dates ISO format YYYY-MM-DD.
   - Principles are declarative, testable, and free of vague language ("should" → 
     replace with MUST/SHOULD rationale where appropriate).

Technique: Template-filling with explicit validation rules, semantic versioning discipline, and cascading consistency propagation (constitution changes must be reflected in plan-template.md, spec-template.md, tasks-template.md, and all command files).


Prompt 3: implement.md (Implementation Gate)

Technique: Prerequisite-checking + checklist-gated execution flow with automatic ignore file management.

2. **Check checklists status** (if FEATURE_DIR/checklists/ exists):
   - For each checklist, count:
     * Completed items: Lines matching `- [X]` or `- [x]`
     * Incomplete items: Lines matching `- [ ]`
   - **If any checklist is incomplete**:
     * Display the table with incomplete item counts
     * **STOP** and ask: "Some checklists are incomplete. Do you want to proceed 
       with implementation anyway? (yes/no)"
     * Wait for user response before continuing

Technique: Programmatic checklist-count logic (regex-based counting of markdown checkbox states) driving a hard stop gate before implementation. This is a novel pattern — the AI is told to count - [X] vs - [ ] patterns as a numeric gate condition.

09

Uniqueness

cline-spec-kit-workflows — Uniqueness

differs_from_seeds

Most similar to spec-kit (Archetype 2 — mirror commands + skills) in spirit, sharing the same artifact sequence (constitution, spec, plan, tasks, implement) and the same GitHub Spec Kit origin. The key architectural difference is the target IDE: spec-kit targets GitHub Copilot while this framework targets Cline. More importantly, this framework relies on GitHub Spec Kit's external Bash scripts (.specify/scripts/) for file management — it is an integration adapter over an existing spec system, not a standalone framework. Compared to openspec (which ships its own script runner and spec format), cline-spec-kit-workflows has no independent runtime. The most distinctive feature is the checklist-gated implementation: the implement workflow counts markdown checkbox states programmatically (- [X] vs - [ ]) and enforces a hard stop if any checklist is incomplete, which is a more rigorous pre-implementation gate than any seed framework.

Positioning

  • Category: Cline-specific adapter layer over GitHub Spec Kit
  • Niche: Teams already using or planning to use GitHub Spec Kit who also use Cline
  • Competitive position: More spec-rigorous than ad-hoc AI coding; requires GitHub Spec Kit as backend

Observable Failure Modes

  1. External dependency risk: The workflows are useless without GitHub Spec Kit's .specify/ scripts. This external coupling is undocumented prominently — users may copy .clinerules and be confused when scripts don't exist.
  2. Script-once constraint: The specify workflow says "You must only ever run this script once" — if the AI re-runs create-new-feature.sh accidentally, it could corrupt the feature directory.
  3. Cascading propagation burden: The constitution workflow instructs updating plan-template.md, spec-template.md, tasks-template.md, and all command files when principles change. This is a heavy manual cascade that is easy to miss.
  4. No automated tests: Despite tracking completion criteria in tasks.md, there is no test execution integrated into the implement workflow.
04

Workflow

cline-spec-kit-workflows — Workflow

Complete Workflow Sequence

[One-Time Setup]
constitution → establishes .specify/memory/constitution.md
    ↓
[Per-Feature]
specify → spec quality check → (clarify?) → plan → (checklist?) → tasks → (analyze?) → implement

Optional phases (marked with ?) can be skipped.

Phase Details

Phase 0: Constitution (One-Time)

Trigger: User runs the constitution workflow once per project. Process:

  1. Load existing .specify/memory/constitution.md (or create if absent)
  2. Collect/derive values for all placeholder tokens
  3. Bump semantic version (MAJOR/MINOR/PATCH based on change type)
  4. Propagate changes to all dependent templates
  5. Produce Sync Impact Report as HTML comment in constitution file Artifact: .specify/memory/constitution.md (updated)

Phase 1: Specify

Trigger: User invokes specify workflow with feature description as $ARGUMENTS. Process:

  1. Run .specify/scripts/bash/create-new-feature.sh → get BRANCH_NAME, SPEC_FILE
  2. Load spec template from .specify/templates/spec-template.md
  3. Extract concepts: actors, actions, data, constraints
  4. Mark ambiguities with [NEEDS CLARIFICATION] (max 3 total)
  5. Write spec to SPEC_FILE
  6. Run inline quality validation → produce FEATURE_DIR/checklists/requirements.md
  7. Auto-repair failed validation items

Approval gate: Up to 3 [NEEDS CLARIFICATION] markers — user must resolve these.

Phase 2: Clarify (Optional)

Trigger: Spec has unresolved [NEEDS CLARIFICATION] markers. Process: Structured questioning to resolve each ambiguity; update spec.

Phase 3: Plan

Trigger: User invokes plan workflow after reviewing spec. Artifact: FEATURE_DIR/plan.md

Phase 4: Checklist (Optional)

Trigger: User wants additional quality validation. Artifact: Checklist files in FEATURE_DIR/checklists/

Phase 5: Tasks

Trigger: User invokes tasks workflow after reviewing plan. Artifact: FEATURE_DIR/tasks.md with completion criteria per task

Phase 6: Analyze (Optional)

Trigger: User wants cross-artifact consistency verification. Process: Reads spec, plan, tasks — identifies contradictions and gaps.

Phase 7: Implement

Trigger: User invokes implement workflow. Process:

  1. Run check-prerequisites.sh to verify required artifacts exist
  2. Check all checklists — if any incomplete, STOP and ask user
  3. Load tasks.md, plan.md, and optional data-model.md, contracts/
  4. Verify/create project ignore files
  5. Execute tasks methodically

Phase-to-Artifact Map

Phase Artifact
Constitution .specify/memory/constitution.md
Specify FEATURE_DIR/spec.md, FEATURE_DIR/checklists/requirements.md
Clarify (updates to spec.md)
Plan FEATURE_DIR/plan.md
Checklist FEATURE_DIR/checklists/*.md
Tasks FEATURE_DIR/tasks.md
Analyze (analysis report, not a file)
Implement modified source files, ignore files

Approval Gates

Gate Type When
[NEEDS CLARIFICATION] markers (max 3) freetext-clarify During specify phase
Incomplete checklist before implement yes-no "Do you want to proceed with implementation anyway?"
Constitution version bump approval freetext-clarify "Propose reasoning before finalizing"
06

Memory Context

cline-spec-kit-workflows — Memory & Context

State Storage

File-based, per-feature-folder pattern. All artifacts written to disk as markdown files.

Memory Files

File Written By Contents
.specify/memory/constitution.md constitution workflow Project principles, governance, versioned
FEATURE_DIR/spec.md specify workflow Feature specification
FEATURE_DIR/checklists/requirements.md specify workflow Spec quality checklist
FEATURE_DIR/plan.md plan workflow Implementation plan
FEATURE_DIR/tasks.md tasks workflow Actionable task list
FEATURE_DIR/checklists/*.md checklist workflow Various quality checklists

Persistence

Project-level persistence via Git-tracked files. All spec artifacts are designed to be version-controlled alongside the code.

Context Loading (implement.md)

The implement workflow explicitly loads multiple context files at the start:

- **REQUIRED**: Read tasks.md for the complete task list
- **REQUIRED**: Read plan.md for tech stack, architecture, and file structure
- **IF EXISTS**: Read data-model.md for entities and relationships
- **IF EXISTS**: Read contracts/ for API specifications
- **IF EXISTS**: Read research.md for technical decisions
- **IF EXISTS**: Read quickstart.md for integration scenarios

This is an explicit multi-file context assembly pattern.

Cross-Session Handoff

Yes (implicit). All artifacts are written to disk, so a new session can re-load them. The implement workflow starts by re-reading tasks.md and plan.md, enabling session continuity.

Compaction

No explicit compaction. Large specs could be problematic in long sessions, but the per-phase design limits what is in context at each stage.

Memory Type Summary

Dimension Value
Memory type file-based
Persistence scope project
State files .specify/memory/constitution.md, FEATURE_DIR/spec.md, FEATURE_DIR/plan.md, FEATURE_DIR/tasks.md
Search mechanism none
Compaction none
07

Orchestration

cline-spec-kit-workflows — Orchestration

Multi-Agent

No. Single-agent, sequential workflow.

Orchestration Pattern

Sequential. Phases are executed in strict order: Constitution → Specify → (Clarify?) → Plan → (Checklist?) → Tasks → (Analyze?) → Implement.

Prompt Chaining

Yes. Each phase's output is the next phase's input:

  • Spec → Plan input
  • Plan → Tasks input
  • Tasks → Implement input
  • Constitution → used by all phases

Isolation Mechanism

None enforced. The workflow creates a feature branch (BRANCH_NAME returned from create-new-feature.sh), but there is no automated git isolation mechanism in the workflow files themselves.

Multi-Model

No. Model-agnostic; no routing.

Execution Mode

Interactive loop. Each phase is invoked by the user as a separate Cline command. There is no continuous execution.

Crash Recovery

Implicit. Since all artifacts are written to disk after each phase, a crash mid-implementation does not lose prior phase artifacts. The user can re-invoke the implement workflow and it will re-read tasks.md.

Consensus Mechanism

None.

Summary Table

Dimension Value
Multi-agent no
Orchestration pattern sequential
Max concurrent agents 1
Isolation none
Consensus none
Prompt chaining yes
Multi-model no
Execution mode interactive-loop
Crash recovery yes (implicit, via disk artifacts)
08

Ui Cli Surface

cline-spec-kit-workflows — UI / CLI Surface

Dedicated CLI Binary

None. No CLI shipped.

Local UI / Dashboard

None.

IDE Integration

Primary: Cline VS Code extension

The .clinerules/workflows/ directory format is how Cline discovers custom workflow commands. Each workflow file becomes a command in Cline's interface, invocable with the file's description as the command name.

Secondary: Any AI assistant that reads markdown

The README states: "The workflows are written in markdown and can be adapted for use with other AI coding assistants that support custom workflows or prompts."

Cline Workflow Command Format

Each workflow file uses YAML front matter that Cline reads:

---
description: Create or update the feature specification from a natural language feature description.
---

The description is shown in Cline's command picker. The $ARGUMENTS token in the workflow body is replaced by text the user types after selecting the command.

Observability

No dedicated observability tools. The artifact files written by each phase serve as an implicit audit trail of what was decided and when.

Cross-Tool Portability

Medium. Cline-specific format (.clinerules/workflows/), but plain markdown content that can be adapted for other tools.

Related frameworks

same archetype · same primary tool · same memory type

Context-Engineering Handbook ★ 9.0k

Provides a first-principles, research-grounded vocabulary and learning path for context engineering — the discipline of designing…

walkinglabs/learn-harness-engineering ★ 6.6k

Teach harness engineering from first principles (12 lectures + 6 projects) and provide a scaffolding skill (harness-creator) that…

Awesome Harness Engineering (walkinglabs) ★ 2.7k

Curate the authoritative reference list of articles, benchmarks, and tools for harness engineering — the practice of shaping the…

cline-memory-bank (nickbaumann98) ★ 581

Custom instructions + 6-file hierarchical Markdown memory bank so Cline maintains full project context across sessions, with a…

FPF (First Principles Framework) ★ 372

Provides a formal pattern language for making reasoning explicit, traceable, and publishable in mixed human/AI engineering work —…

nexu-io/harness-engineering-guide ★ 134

Provide a practical, code-first reference guide to harness engineering — from first principles to production patterns —…