Agent skill

Spec Driven Development

by kdlbs in kdlbs/kandev

Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification.

AGPL-3.0Auto-check passedDevelopment

Install Spec Driven Development

skills CLI
$ npx skills add kdlbs/kandev --skill spec-driven-development -a claude-code

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

GitHub CLI
$ gh skill install kdlbs/kandev spec-driven-development --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/kdlbs/kandev.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/spec-driven-development .claude/skills/spec-driven-development && 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
spec-driven-development
GitHub stars
909
Token cost
~1.9k tokens
SKILL.md length
1,024 words
Files
1
Skills in repo
45
Repo updated
First seen
Licence
AGPL-3.0

At a glance

Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification.

  • Works in 3 steps: Track the work → Execute work orders → Open the PR
  • Tasks that involve Spec-driven development
  • SKILL.md covers Core flow, Prompt-precedence gate, Phase 0: Track the work and Phases 1 to 4: Design, plus 5 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Spec Driven Development is an agent skill from kdlbs/kandev. Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Development, covering Spec-driven development. The repository describes itself as: AI Kanban & Development Environment. Orchestrate multiple agents, review changes, open PRs. Multi-provider, self-hostable, no telemetry. The licence is AGPL-3.0.

When your agent uses it

  • Tasks that involve Spec-driven development

Example prompts

  • “/spec-driven-development”

Workflow steps

3 steps, taken from the step headings in SKILL.md.

  1. Track the work
  2. Execute work orders
  3. Open the PR

What it can do on your machine

Read from SKILL.md and the folder at commit b734113. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

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

  • Network

    No URLs in SKILL.md.

    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.

Context cost

Spec Driven Development loads about 1.9k tokens when it runs. Until then it costs about 51 tokens; SKILL.md has 1,024 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~51
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k

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 kdlbs/kandev at commit b734113, republished under its AGPL-3.0 licence (© kdlbs). 1,024 words, ~1,943 tokens.

Download SKILL.mdSave it as .claude/skills/spec-driven-development/SKILL.md (or your agent's skills folder).
name
spec-driven-development
description
Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification.

Spec-Driven Development

Use this workflow for non-trivial features and behavior-changing fixes. The user-started primary conversation performs every phase by default. Subagents require explicit user authorization.

Core flow

text
Intent -> Requirements -> System design -> Plan + work orders -> Design-package handoff -> Explicit implementation request -> TDD implementation -> PR review -> Report

Do not replace durable artifacts with chat-only notes. Do not skip from intent to code unless the user explicitly opts out.

Prompt-precedence gate

Before you edit production or permanent test files, make sure that one condition is true:

  1. A prior design turn produced the requirements, system design, plan, and current work order. The user now explicitly asks to implement them.
  2. The request references existing reviewed specifications, a plan, and the current work order.
  3. The user explicitly says to skip specification, plan, and work-order creation.

Workflow envelopes, phase labels, TDD lists, and generic implementation prompts do not satisfy this gate.

If no condition is true, create or update the design package. Record any partial implementation as continuation context. Stop at the design-package handoff.

When the user asks for feature planning, complete the full design package before you return control. Stop after requirements only when the user requests a requirements review or a material question blocks safe design.

Phase 0: Track the work

Keep a visible task list:

  1. Clarify intent.
  2. Create or update requirements.
  3. Create or update the system design.
  4. Create the plan and work orders.
  5. End the design turn.
  6. Execute work orders with TDD after an explicit implementation request.
  7. Open the PR, address valid findings, record changes, and report.

When task-plan MCP tools are available, use the Kandev task plan for live phase progress and update it after each completed design phase. Keep docs/plans/<initiative>/plan.md and its work orders as the durable implementation package; neither plan replaces the other.

Phases 1 to 4: Design

Run the /interview-me assumption check before drafting the package. Reuse confirmed choices across phases; ask only about material unresolved decisions. For large uncertain initiatives, use its optional decision-mapping reference before creating detailed work orders. Continue the package once blockers resolve.

Run the /spec ownership gate before you choose paths. Search adjacent systems for the same capability. Choose the owner from the durable contract, not from backend or frontend implementation layers. If a feature has a UI, include its observable UI outcomes and frontend design in the feature owner's artifacts. Create a separate UI artifact only for an independent reusable UI contract.

Create or update:

  • docs/specs/<system>/requirements/<capability>.md through /spec.
  • docs/specs/<system>/system-design/<capability>.md through /spec.
  • docs/plans/<initiative>/plan.md through /plan.
  • docs/plans/<initiative>/task-<NN>-<short-slug>.md through /plan.

During migration, use a legacy specification only when no new requirement or system design replaces it.

Requirements define stable REQ-* and AC-* IDs. System designs map technical behavior to those IDs. Plans define delivery order. Work orders define one implementation result and its verification.

Keep one vertical requirement/design pair for one cohesive outcome. Work orders can separate implementation boundaries after the specification has one owner.

Each work order must contain:

  • Frontmatter with id, title, status, wave, depends_on, plan, requirements, acceptance_criteria, and system_design.
  • A summary, scope, and exclusions.
  • One to three implementation acceptance conditions.
  • Exact targeted verification commands.
  • Specific likely files, dependencies, and risks.

Work orders do not name a worker role or model tier. Waves can identify parallel-safe candidates. They do not authorize subagents.

For rendered UI changes, include ASCII previews in the plan and relevant UI work orders through /plan. Follow the shared contract in docs/specs/guide/plans-and-work-orders.md#ascii-ui-previews.

Design-package handoff

Before implementation, report:

  • Requirement documents and IDs.
  • System-design paths.
  • The plan and work-order paths.
  • Dependency order and exact verification commands.
  • Open risks and exclusions.
  • For UI changes, a compact inline ASCII preview of the proposed UI, with a distinct phone view when composition differs, plus links to the full previews.

Then end the turn. Do not ask the user to approve the package or switch models. The user reviews the files and sends a later implementation request.

Recommended Codex route:

  • GPT-5.6 Sol/high for intent, investigation, specifications, plans, and high-risk design.
  • GPT-5.6 Terra/medium for implementation, TDD, and integration.
  • GPT-5.6 Luna/low for short and mechanical read-only work.

The user controls model selection. Do not infer a model change from prose.

Show full SKILL.md (352 more words)Show less

Phase 5: Execute work orders

For each work order, in dependency order:

  1. Read the work order, requirements, system design, plan, and scoped AGENTS.md.
  2. Change only that work order to in_progress.
  3. Implement with /tdd. Use /e2e and /mobile-parity when applicable.
  4. Run the exact targeted verification commands.
  5. After implementation and targeted checks pass, change only the current work order to done.
  6. After every linked work order passes, verify the implementation still matches each paired requirement and system design. If it does, promote each paired draft requirement to active and each paired draft system design to current, then change the plan to implemented. Record spec lifecycle in spec frontmatter and delivery status in plans and work orders; docs/specs/INDEX.md is a static entry page, not a per-spec delivery catalog. Update a system README only when its boundary or migration record changed. If implementation diverged or an open design issue remains, do not promote unresolved draft specs or mark the plan implemented; report the mismatch.

If the user authorizes subagents, launch only parallel-safe work orders in the requested wave. Use native delegation with no full-history fork. Give each child the work-order path, owned files, dependencies, and exact command.

If work needs a new architecture, public contract, persistence boundary, or high-impact security decision, stop and request a strong-model design pass. Use /record for a durable decision.

Phase 6: Open the PR

After all work-order checks pass, request explicit user authorization before running /commit, /push, or /pr. Run only the operations the user authorizes. Do not add automatic local simplify, QA, broad review, security review, or /verify work.

The configured PR reviewers are the semantic-review gate. Use /pr-fixup only for a CI error or an actionable reviewer finding. Run the affected work-order checks after a correction.

Stop conditions

Stop and ask the user when specifications and code disagree, implementation needs a material new design, a model escalation is necessary, or the same check fails after three focused attempts.

Final report

Report requirement IDs, system-design and plan paths, work-order statuses, model checkpoints, changed files, commands and results, user-authorized subagents, PR-review evidence, and known risks.

© kdlbs, AGPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/spec-driven-development of kdlbs/kandev.

Open the folder on GitHubat commit b734113

Compare with similar skills

Spec Driven Development 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.

Spec Driven Development compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Spec Driven Development this skillkdlbs/kandev909—~1.9kAutomated safety check: PassAGPL-3.0
OpenSpec Bulk Change ArchiverFission-AI/OpenSpec71k3 repos~5.6kAutomated safety check: PassMIT
Speckit ConstitutionWeihanLi/WeihanLi.Common24211 repos~2.1kAutomated safety check: PassApache-2.0
Speckit Taskstoissueskunstmusik/blue15419 repos~2kAutomated safety check: PassGPL-3.0
Speckit Analyzekunstmusik/blue15418 repos~3kAutomated safety check: PassGPL-3.0
Review Spdzhu1090093659/spec_driven_develop984—~1.5kAutomated safety check: PassMIT

Similar skills

  • Archives several completed OpenSpec changes in one operation, checking the codebase to resolve spec conflicts rather than archiving blindly.

    71k GitHub starsUsed in 3 repos~5.6k tokens
    DevelopmentAuto-check passed
  • Speckit Constitution

    WeihanLi/WeihanLi.Common

    Create or update the project constitution from interactive or provided principle inputs, ensuring all dependent templates stay in sync.

    242 GitHub starsUsed in 11 repos~2.1k tokens
    DevelopmentAuto-check passed
  • Speckit Taskstoissues

    kunstmusik/blue

    Convert existing tasks into actionable, dependency-ordered GitHub issues for the feature based on available design artifacts.

    154 GitHub starsUsed in 19 repos~2k tokens
    DevelopmentAuto-check passed
  • Speckit Analyze

    kunstmusik/blue

    Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.

    154 GitHub starsUsed in 18 repos~3k tokens
    DevelopmentAuto-check passed
  • Review Spd

    zhu1090093659/spec_driven_develop

    Findings-first code review workflow for AI coding agents. An agent skill from zhu1090093659/spec_driven_develop.

    984 GitHub stars~1.5k tokensUpdated 2 mo ago
    DevelopmentAuto-check passed
  • Speckit Plan

    kunstmusik/blue

    Execute the implementation planning workflow using the plan template to generate design artifacts.

    154 GitHub starsUsed in 18 repos~2.1k tokens
    DevelopmentAuto-check passed

More from kdlbs/kandev

All 45 skills in this repo
  • PR Walkthrough

    kdlbs/kandev

    Generate a single-file HTML walkthrough that explains a PR's purpose, user impact, interface changes, compatibility risks, and implementation.

    909 GitHub stars~6.2k tokensUpdated today
    Auto-check passed
  • Debug

    kdlbs/kandev

    Diagnose Kandev bugs, running-instance issues, UI/browser failures, and runtime behavior.

    909 GitHub stars~2.2k tokensUpdated today
    Auto-check passed
  • Improve Kandev's AI harness from session learnings or explicit requests.

    909 GitHub stars~1.3k tokensUpdated today
    Auto-check passed
  • Diagram Design

    kdlbs/kandev

    Create branded architecture, IT current-state, flowchart, sequence, state machine, ER/data model, timeline, swimlane, quadrant, radar/spider, polar chart (polar/radial lollipop), loop/flywheel…

    909 GitHub starsUsed in 1 repo~8k tokens
    Auto-check passed
  • TDD

    kdlbs/kandev

    Implement changes using Test-Driven Development (Red-Green-Refactor).

    909 GitHub stars~4.2k tokensUpdated today
    Auto-check passed
  • Verify

    kdlbs/kandev

    Run a broad local verification audit only when the user explicitly requests it or PR/CI remediation requires it.

    909 GitHub stars~2.7k tokensUpdated today
    Auto-check passed

Categories

Questions about Spec Driven Development

What does Spec Driven Development do?

Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification. Spec Driven Development is an agent skill from kdlbs/kandev. Single-session Kandev feature workflow: clarify intent, create durable requirements, system designs, plans, and work orders, then hand off before implementation and verification.

When should I use Spec Driven Development?

Spec Driven Development fits situations like: tasks that involve Spec-driven development.

How do I install Spec Driven Development in Claude Code?

Run `npx skills add kdlbs/kandev --skill spec-driven-development -a claude-code`. Or copy the skill folder (.agents/skills/spec-driven-development in kdlbs/kandev) into .claude/skills/spec-driven-development in your project. Claude Code loads it when a task matches its description.

How do I install Spec Driven Development in Codex?

Run `npx skills add kdlbs/kandev --skill spec-driven-development -a codex`. Or copy the skill folder (.agents/skills/spec-driven-development in kdlbs/kandev) into .agents/skills/spec-driven-development in your project. Codex loads it when a task matches its description.

Can I use Spec Driven Development 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 kdlbs/kandev --skill spec-driven-development -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/spec-driven-development, .gemini/skills/spec-driven-development, .github/skills/spec-driven-development and .opencode/skills/spec-driven-development in your project.

What does Spec Driven Development need to run?

SKILL.md names no scripts, command-line tools or credentials: Spec Driven Development is instructions for the agent only.

Does Spec Driven Development access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Spec Driven Development 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 Spec Driven Development use?

Spec Driven Development is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Spec Driven Development use?

About 1.9k tokens (SKILL.md is roughly 7.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Spec Driven Development?

Skills that share tags, products or a category with Spec Driven Development: OpenSpec Bulk Change Archiver (Fission-AI/OpenSpec, 71k stars), Speckit Constitution (WeihanLi/WeihanLi.Common, 242 stars), Speckit Taskstoissues (kunstmusik/blue, 154 stars) and Speckit Analyze (kunstmusik/blue, 154 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Spec Driven Development?

kdlbs (a GitHub organization) maintains it in kdlbs/kandev, which has 909 GitHub stars. The repository holds 45 skills in this directory. The repository was last updated on October 8, 2026.

Source: kdlbs/kandev on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.