A semi-automated pipeline for taking one software task, called a phase, from discussion through planning, implementation, verification, commit, and pull request creation. TDD means writing a failing test before the code that makes it pass.

Use it to start or continue a phase, review its plan, implement it with or without TDD, verify the result, commit and push changes, and create a pull request.

It keeps the work moving through defined review and implementation stages, with TDD enabled by default and an option to skip it.

pipeline — Skill for Claude Code by seungdeok

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add seungdeok/harness-starter --skill pipeline

Clone the repo

git clone --depth 1 https://github.com/seungdeok/harness-starter

Made for: Claude Code.

Or install harness, the plugin that ships this one along with the rest of its 6 skills, 1 hook.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/seungdeok/harness-starter/pipeline/github.svg)](https://agentmods.dev/skills/seungdeok/harness-starter/pipeline)

Your own site

<a href="https://agentmods.dev/skills/seungdeok/harness-starter/pipeline"><img src="https://agentmods.dev/badge/skills/seungdeok/harness-starter/pipeline/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/seungdeok/harness-starter/pipeline"><img src="https://agentmods.dev/badge/skills/seungdeok/harness-starter/pipeline.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 176 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,094 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00176	$0.03094
Opus 5	$0.00088	$0.01547
Sonnet 5	$0.00035	$0.00619
Haiku 4.5	$0.00018	$0.00309

Measured 9d ago against content hash 4b9857c8cc18, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

pipeline scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/pipeline.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/pipeline/SKILL.md · 168 lines

How it starts

The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.

pipeline

pipeline.py 를 몰아서 한 phase 를 discuss 부터 make-pr 까지 진행해요. compound(CLAUDE.md 5장)는 안 해요 — 파이프라인이 끝나면 사용자가 /ce-compound 로 직접 남겨요.

Stages: discuss → plan → [plan-review-ceo] → [plan-review-eng] → approve → implement-red → implement-green (TDD, 기본) | implement (--no-tdd) → verify → commit-push → make-pr

스크립트 위치 (먼저 결정)

아래 규칙으로 pipeline.py 경로를 정하고, 이후 모든 명령에서 그 경로를 써요:

환경변수 HARNESS_SCOPE 를 확인하고, 없으면 프로젝트의 .claude/settings.local.json 의 env.HARNESS_SCOPE 를 읽어요. 값이 project 이고 <프로젝트>/scripts/pipeline.py 가 있으면 → 그 파일을 사용.
그 외(글로벌 scope, 또는 값 없음) → 이 스킬 폴더의 scripts/pipeline.py(번들) 를 절대 경로로 사용.
HARNESS_SCOPE 값이 어디에도 없으면 /harness:setup 을 먼저 실행하라고 권한 뒤, 사용자가 원하면 번들 스크립트 + 기본값으로 그냥 진행해도 돼요.

스크립트는 cwd 기준 git root 에 phase 를 만들므로, 어느 복사본이든 실행 위치(cwd)가 곧 대상 레포예요.

0. 시작 전 점검과 질문 (필수)

0-1. 하드 의존 스킬 점검

init 은 worktree·브랜치를 먼저 만들기 때문에, 스킬 누락을 stage 도달 시점에 발견하면 이미 늦어요. init 전에 사용 가능한 스킬 목록에서 아래를 확인해요. 목록에 뜨는 식별자 그대로 찾아요 — OMC 는 oh-my-claudecode: 접두어가 붙고, gstack 은 프로젝트 .claude/skills/ 에 있어 접두어가 없어요.

stage	스킬	목록상 이름	없을 때
plan	`/plan`	`oh-my-claudecode:plan`	대안 없음
implement(-red/-green)	`/ultrawork`	`oh-my-claudecode:ultrawork`	대안 없음
verify	`/verify`	`oh-my-claudecode:verify`	대안 없음
plan-review-ceo	`/plan-ceo-review`	`plan-ceo-review`	`--no-review`
plan-review-eng	`/plan-eng-review`	`plan-eng-review`	`--no-review`

OMC 스킬이 하나라도 없으면 어느 stage 가 막히는지 알리고, AskUserQuestion 으로 설치 후 다시 시작 / 그대로 진행 을 물어요. 임의로 init 을 강행하지 않아요.
gstack 스킬이 없으면 아래 plan review 질문을 묻지 않고 --no-review 로 고정해요. 고를 수 없는 걸 선택지에 남기지 않아요.

0-2. 시작 전 질문

init 하기 전에 AskUserQuestion 으로 세 가지를 먼저 물어요.

worktree — 이 phase 를 어디서 돌릴지
- 전용 worktree (기본, 병렬 안전): 레포 형제 디렉토리 ../<repo>-worktrees/<slug> 에 새 브랜치로 격리 → init 에 플래그 없음
- 현재 체크아웃: 지금 브랜치에서 바로 → init --no-worktree
- 위치를 바꾸려면 init --worktree-dir <부모경로> — <slug> 이 뒤에 붙고 상대 경로는 메인 체크아웃 기준이에요. 예전 위치를 그대로 쓰려면 --worktree-dir .claude/worktrees. (--no-worktree 와 동시에 줄 수 없어요.)
plan review — 계획 검토를 어디까지 할지
- CEO + Eng 둘 다 (기본) → 플래그 없음
- CEO 만 / Eng 만 → 둘 다 넣되(플래그 없음) 원치 않는 review stage 에서 실행 없이 advance
- 생략 → init --no-review
TDD — implement 를 RED/GREEN 두 stage 로 나눌지
- yes (기본) → 플래그 없음
- no → init 에 --no-tdd 를 늘 명시적으로 전달

Read the full file on GitHub · 168 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

scripts/pipeline.py 33 KB runs code

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

9d ago First seen · 168 lines · 176 tokens per session scan A 4b9857c8cc18

Subscribe to this mod's changes

pipeline is a skill published in the GitHub repository seungdeok/harness-starter (2 stars, last pushed 23d ago), licensed MIT. It adds 176 tokens to every session and 3,094 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Planning TDD Testing

pipeline

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

pipeline

스크립트 위치 (먼저 결정)

0. 시작 전 점검과 질문 (필수)

0-1. 하드 의존 스킬 점검

0-2. 시작 전 질문

What ships with it

What this file has done since we first saw it

Other skills, from other repositories

engram-testing-coverage

iterative-development

python

nw-fp-clojure

strict-tdd

mobiai-mobile-tdd