harness-aijient-team

Plugin harness-aijient-team from marketplace harness-aijient-team-marketplace in bd-makers/team-harness. Commands read from the repo's manifest — 6 days ago.

Skills (25)

• harness-codex-sim

  Run Codex headless L5 harness simulations with `tests/sim/codex-agentloop.mjs`. Use when verifying the Harness AIjient Team Codex plugin, `codex exec` skill activation, JSONL evidence, installed harness task/init flows, or Codex-specific PASS/FAIL/MANUAL/N/A reports.

• harness-release

  Codex wrapper for releasing the Harness AIjient Team plugin. Use when the user asks for /harness-release, harness release, manifest version bump, marketplace sync, or plugin cache sync.

• harness-sim

  설치된 하네스와 에이전트 워크플로우 스킬(`/harness-unittest`·`/harness-comptest`)이 진짜 작동하는지 실제 claude -p 에이전트 세션으로 검증하고(L5 agent-in-the-loop) 날짜 리포트를 남긴다. "하네스 시뮬레이션", "harness sim", "playground 검증", "하네스 동작 점검", "설치된 하네스 테스트", "스킬 검증(skilltest)" 요청에 사용.

• harness-adversarial-review

  Engine-neutral wrapper for the adversarial variant of the harness review - it challenges implementation approach and design choices of local git changes with any engine (codex, claude, custom). Use when the user asks for /harness-adversarial-review, adversarial review, or a review that tries to reje

• harness-comptest

  Codex wrapper for writing Testing Trophy component tests in React/React Native projects. Use when the user asks for /harness-comptest, component or screen test writing, Testing Library integration tests, user-event interaction tests, form-flow tests, msw-boundary component tests, or React Native Tes

• harness-contrarian

  Codex wrapper for challenging active task assumptions. Use when the user asks for /harness-contrarian, contrarian review, assumption challenge, hidden cost review, or plan/spec pushback.

• harness-diagram

  Harness adapter for the active task's spec/plan diagram — generates or refreshes the one diagram artifact for the current harness task under the docs user task directory, enforcing the harness conventions (self-contained inline SVG, generated-not-SSOT status, probe/degrade/record, artifact record).

• harness-doctor

  Codex wrapper for checking Harness AIjient Team install health. Use when the user asks for /harness-doctor, harness doctor, harness health check, symlink check, hook check, or settings validation.

• harness-init

  Codex wrapper for scaffolding the Harness AIjient Team workflow into the current project. Use when the user asks for /harness-init, harness init, initialize team harness, or scaffold AGENTS.md workflow.

• harness-interview

  Codex wrapper for Socratic spec clarification on the active task. Use when the user asks for /harness-interview, harness interview, ambiguity review, spec interview, or Socratic task questions.

• harness-inttest

  Codex wrapper for writing Khorikov Part III integration tests in JS/TS backend projects. Use when the user asks for /harness-inttest, integration test writing, API route or handler tests, repository or database access tests, testcontainers setup, msw-node outbound-HTTP boundary tests, or process-bou

• harness-loop

  Codex wrapper for the optional default implementation loop. Use when the user asks for /harness-loop, harness loop, run the plan with an orchestrator, or a Dev → QA → commit loop over the active task plan.

• harness-migrate

  Codex wrapper for migrating older harness layouts to the current structure. Use when the user asks for /harness-migrate, harness migrate, backup script migration, or old task structure migration.

• harness-observe

  Codex wrapper for the harness observability scorecard and trip wires. Use when the user asks for /harness-observe, harness observe, tool failure trends, trip wire status, or observability logs.

• harness-promote

  Codex wrapper for promoting a task artifact Learnings entry into .claude/rules with a provenance marker. Use when the user asks for /harness-promote, harness promote, rules promote, or to turn a repeated learning into a project rule.

• harness-retro

  Codex wrapper for appending learnings to the active task artifact. Use when the user asks for /harness-retro, harness retro, record a learning, task retrospective, or self-improvement note.

• harness-review

  Engine-neutral wrapper for running the read-only external review (codex, claude, or custom engine) of local git state and recording the result in the active task artifact. Use when the user asks for /harness-review, harness review, external code review, or a review of current changes with a specific

• harness-ship

  Codex wrapper for the pre-PR final documentation pass. Use when the user asks for /harness-ship, harness ship, prepare a PR or MR, or a final spec/plan/artifact update before opening a pull request.

• harness-simplifier

  Codex wrapper for simplifying the active task plan. Use when the user asks for /harness-simplifier, harness simplifier, YAGNI review, remove overengineering, or simplify plan steps.

• harness-spec

  Codex wrapper for drafting the active task spec from Confluence, Figma, or an interview. Use when the user asks for /harness-spec, harness spec, spec draft, spec generation, or writing spec.md from PRD or design sources.

• harness-sync

  Codex wrapper for regenerating the .cursor/rules mirror from .claude/rules and reinstalling the git post-commit hook. Use when the user asks for /harness-sync, harness sync, or to sync Claude and Cursor rules.

• harness-task

  Codex wrapper for managing Harness AIjient Team task docs. Use when the user asks for /harness-task, harness task, create task, activate task, list tasks, or complete task docs.

• harness-team

  Use the harness-team project workflow from Codex. Trigger when the user asks to initialize, refresh, inspect, repair, release, or operate the Team Harness; create or resume task docs under the docs user task structure; run harness doctor; or work inside a project configured by the Claude Code Team H

• harness-unittest

  Codex wrapper for writing Khorikov-style unit tests in JS/TS/React/React Native projects. Use when the user asks for /harness-unittest, unit test writing, test coverage for a file or feature, Given-When-Then tests, Testing Library tests, or Vitest/Jest test authoring.

• harness-wiki

  Codex wrapper for the optional wiki compile. Use when the user asks for /harness-wiki, harness wiki, compile a merged task into the wiki, or turn task decisions and learnings into feature or module wiki entries with PR provenance.

Slash commands (21)

• /harness-aijient-team:harness-adversarial-review

  설계 결정과 가정에 반박을 시도하는 적대적 read-only 외부 리뷰 — 절차·엔진 표는 harness-review와 동일, 리뷰 프롬프트만 교체

• /harness-aijient-team:harness-comptest

  React/React Native 컴포넌트·화면·UI 플로우에 Testing Trophy 통합 층 테스트를 작성한다 (스택 감지 → 사용자 관점 검증 강제 → 커버리지 → 검증)

• /harness-aijient-team:harness-contrarian

  활성 task의 spec.md / plan.md의 모든 가정에 의문을 제기한다 — 엔진 인수를 주면 외부 read-only 검증자가 반론(D6)

• /harness-aijient-team:harness-diagram

  활성 task의 spec/plan 다이어그램을 하네스 규약(경로·자립형 inline SVG·artifact 기록)대로 생성하거나 갱신

• /harness-aijient-team:harness-doctor

  하네스 무결성 점검 (파일, hooks, CLI PATH)

• /harness-aijient-team:harness-init

  현재 프로젝트에 팀 하네스를 scaffold하거나 기존 설치를 갱신합니다 (Claude 메인 + Codex 리뷰어 + Cursor 미러). 마커 병합·JSON deep-merge라 재실행해도 사용자 텍스트를 보존

• /harness-aijient-team:harness-interview

  활성 task의 spec.md를 선행 채점 후 소크라테스식 질문으로 검증해 숨겨진 가정을 드러낸다

• /harness-aijient-team:harness-inttest

  API 핸들러·DB 접근·캐시·아웃바운드 HTTP 등 프로세스 경계 코드에 Khorikov Part III 통합 테스트를 작성한다 (스택 감지 → managed/unmanaged 구분 강제 → 커버리지 → 검증)

• /harness-aijient-team:harness-loop

  plan 확정 뒤 선택형 기본 루프 — 메인 세션이 오케스트레이터가 되어 plan 단계마다 Dev 구현 → QA 판정 → 커밋을 순차로 돌리고, 네 멈춤 조건에서 사람에게 넘긴다

• /harness-aijient-team:harness-migrate

  구버전 구조를 최신으로 마이그레이션 + 설치된 훅·스킬·규칙을 최신 템플릿으로 refresh — init은 기존 파일을 건너뛰므로 이 파일들의 템플릿 수정이 도달하는 유일한 경로 (AGENTS.md·CLAUDE.md 관리 절은 init이 렌더한다)

• /harness-aijient-team:harness-observe

  관측 로그(.harness/observability) 스코어카드와 트립와이어 판정 — read-only, 발화 시 exit 1

• /harness-aijient-team:harness-promote

  활성 task artifact의 Learnings 항목을 사용자 승인 후 .claude/rules/<slug>.md로 승격 — 유래 마커 부착·cursor 미러 재생성

• /harness-aijient-team:harness-retro

  현재(또는 지정) task의 artifact.md에 학습/교정 내용을 append. CLAUDE.md 자기개선 루프 정책 실행.

• /harness-aijient-team:harness-review

  엔진 중립 read-only 외부 리뷰(codex·claude·custom)를 로컬 git 상태에 실행하고 결과를 활성 task의 artifact 'Reviews' 절에 기록

• /harness-aijient-team:harness-ship

  PR/MR 직전 spec·plan·artifact를 최종 갱신하고 다이어그램(옵트인)까지 정리한 뒤 준비 완료 상태를 보고

• /harness-aijient-team:harness-simplifier

  활성 task의 plan.md에서 제거 가능한 단계와 추상화를 찾아낸다 — 엔진 인수를 주면 외부 read-only 검증자가 채점(D6)

• /harness-aijient-team:harness-spec

  활성 task의 spec.md 초안을 Confluence·Figma·인터뷰 3소스에서 생성한다 (writer — 검증은 /harness-interview)

• /harness-aijient-team:harness-sync

  .claude/rules → .cursor/rules 미러 재생성(단방향) + post-commit 훅 재설치

• /harness-aijient-team:harness-task

  task 관리 (task/list/done/handoff) — docs/<user>/<name>/ 구조

• /harness-aijient-team:harness-unittest

  JS/TS/React/RN 프로젝트에 Khorikov 원칙 기반 단위테스트를 작성한다 (스택 자동 감지 → GWT 강제 → 커버리지 → 검증)

• /harness-aijient-team:harness-wiki

  머지된 task의 결정·바뀐 모듈·학습을 PR·커밋·작성자 출처가 붙은 기능·모듈 위키 항목(wiki/)으로 컴파일 — 선택 단계, 머지 후 종결 절차에서 done 다음

Similar plugins