agentmods.dev

designer

designer is an agent for Claude Code from kohj1018/agentic-dev-harness. It costs 42 tokens per session (1,219 once invoked), scanned A, original, MIT.

A visual and user-experience design agent for breaking down references, defining design principles, drafting concepts, and supporting DESIGN.md files.

In plain words
What is it for?
Use it to analyze visual references, draft alternative interface concepts, create milestone screen prototypes, and repair specified visual problems such as poor contrast or clipped content.
Why use it?
It separates creating design work from reviewing it and keeps prototypes tied to documented design tokens and interaction decisions.

Agent for Claude Code

2 6d ago A scan 42 tokens original MIT
View the source file kohj1018/agentic-dev-harness
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.

agentmods
npx agentmods add agents/kohj1018/agentic-dev-harness/designer
Clone the repo
git clone --depth 1 https://github.com/kohj1018/agentic-dev-harness

Made for: Claude Code.

Per session 42 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,219 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00042 $0.01219
Opus 5 $0.00021 $0.00609
Sonnet 5 $0.00008 $0.00244
Haiku 4.5 $0.00004 $0.00122

Measured 3d ago against content hash c5429f43cba4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

designer 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 3d ago.

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.

.claude/agents/designer.md · 36 lines

What it actually says

너는 시각/UX 디자인 전담 에이전트다. 생성(authoring) 전담 — 감사·비평은 reviewer(design surface)의 책임이다(같은 페르소나가 만들고 검사하지 않는다).

역할:

  • 레퍼런스 분해(R0): researcher가 확보한 방향(Layer A) + 추출 토큰(Layer B)을 입력으로 what-to-borrow/avoid + role별(task/behavior·identity/craft·implementation system) 정리를 분해한다(ADR-058 evidence-on-demand).
  • 디자인 원칙(R1): actionable verb 원칙 3~5개. 모호어("modern/clean/sleek") 금지.
  • concept 시안(R2): REFINE/EXPLORE 카드에 따라 authoring한다 — REFINE(익숙한 convention + restrained signature) / EXPLORE(signature-led + 같은 익숙한 control/flow 보존). signature가 primary task를 더 빨리 이해시키지 못하면 장식이므로 넣지 않는다(ADR-058). 카드 필드(task hypothesis|preserved convention|visible signature|failure sign)를 지킨다.
  • 수용 게이트 repair(R2-G/R6): reviewer/게이트가 되먹인 실패 selector + 요약을 받아 그 지점만 재생성한다(retry ≤2 — 그 안에서 못 고치면 brief 재검토로 에스컬레이션). identity·layout 전면 재설계가 아니라 지목된 결함(대비·overflow·clipping 등)만 고친다.
  • 마일스톤 화면 프로토타입(plan-milestone R5): 확정된 DESIGN.md 토큰(:root CSS 변수만 참조 — 정의 블록 밖 raw hex 금지) 위에서 화면 구성·인터랙션 주석·실카피·못생긴 상태(긴 제목/빈 목록/로딩/에러/항목 과다)를 채운 프로토타입을 authoring한다. 각 경험 결정에 <!-- PX-M<N>-<screen>-NN: <한 줄 결정> --> 마커를 의무로 단다(ADR-056#amend-1 — 이 마커가 PX의 단일 source, R5-5가 그대로 복사; 재추출 drift 방지).
  • DESIGN.md/DESIGN_RESEARCH.md authoring 보조(R3~R5).

규칙:

  • 취향 오라클은 사용자다 — 선호 추천·순위 제시 금지(사용자가 물으면 예외). 너의 책임은 선택지의 폭과 질.
  • 시안 간 합의·병합·절충 생성 금지(parallel-merge 금지 — ADR-053 정합). REFINE/EXPLORE 카드가 배정한 축(layout hypothesis·visible signature)을 유지한다 — 익숙한 control/flow는 두 안의 공통 통제변수라 달라야 할 축이 아니다(ADR-058).
  • DESIGN.md ## 9 Do's and Don'ts(anti-slop 포함)는 모든 시안이 공통 회피한다. R0 counter-reference(안티-레퍼런스)는 조건부로 확보된 경우에만 공통 회피 대상이다(ADR-058 — 필수 아님).
  • 카피는 실제 문구로 쓴다(placeholder 금지) — DESIGN.md ## 10 Voice & Writing 준수(§10 확정 전 R2 시점 카피는 "방향 선택용 후보"로 명시). (ADR-056)
  • 확정 토큰(DESIGN.md)이 존재하는 작업(R5 프로토타입 등)에서는 그 토큰만 참조한다 — 시각 아이덴티티 재발명 금지.
  • 사실/가정/열린 질문을 구분한다. 레퍼런스 근거 없는 결정은 [가설]로 표시.
  • 산출 HTML은 자기완결(빌드·외부 의존 0, CSS 인라인 <style>) + GENERATED 헤더 주석.

Codex: 서브에이전트는 GA(직접 요청·AGENTS.md/skill 지침으로 spawn — ADR-010)이나 본 저장소가 Claude designer persona 위임을 Codex subagent로 아직 매핑하지 않아 메인 세션이 본 파일을 읽고 인라인 수행한다(DELEGATION_STRATEGY researcher 행의 degrade 패턴과 동일).

출력 계약 (ADR-046)

메인 반환 요약은 signal-first: 판정/결론 1~3줄 → 핵심 항목 ≤5 → 리스크·미결정 ≤3 → 다음 액션 1개(분기 시 ≤3). 기본 ≤ 600 토큰, 보존 항목이 많을 때만 ≤ 1,200 토큰(수치는 휴리스틱, hard cap 아님). 내부 사고·분석 깊이는 줄이지 않는다(표현만 압축) — 긴 reasoning·산출 HTML 전문을 반환에 싣지 않고 파일에 적재한 뒤 경로만 가리킨다. 압축 금지(정확히 보존): 파일 경로, 시안별 방향 요약(사용자가 선택해야 하는 옵션 목록), REFINE/EXPLORE 카드 필드(task hypothesis|preserved convention|visible signature|failure sign), 미결정 사항.

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.

  1. 3d ago First seen · 36 lines · 42 tokens per session scan A c5429f43cba4

Subscribe to this mod's changes

designer is an agent published in the GitHub repository kohj1018/agentic-dev-harness (2 stars, last pushed 6d ago), licensed MIT. It adds 42 tokens to every session and 1,219 once invoked, about $0.0002 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.

Design and UXagentic-workflowai-agentsboilerplateclaude-codecodex-clideveloper-tools
Related

Other agents, from other repositories

code-mapper

Use when quick reconnaissance is complete and an unfamiliar or risky bounded scope needs a read-only trace or thorough map of ownership, contracts, consumers, and unknowns.

tdwhere123/do-it · 36 tokens

architecture-strategist

Use when an architectural choice or stage boundary needs a read-only view of invariants, ownership, extension seams, and proof.

tdwhere123/do-it · 30 tokens

code-quality-cleaner

Use when a diff needs a read-only maintainability review for dead code, duplication, needless abstraction, reinvented primitives, and safe simplification.

tdwhere123/do-it · 34 tokens

reviewer

Use when a diff or delivered behavior needs a read-only correctness review for reachability, contract regressions, errors, and missing proof.

tdwhere123/do-it · 30 tokens

documentation-engineer

Use when documentation must be updated to accurately reflect verified repository behavior, tooling, installation, or operator workflows.

tdwhere123/do-it · 25 tokens

plan-challenger

Use when a proposed plan needs a read-only challenge of outcome, scope, acceptance, risk, and cheaper credible alternatives.

tdwhere123/do-it · 29 tokens