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.
npx agentmods add agents/byeongminlee/nextjs-claude-code/task-executorgit clone --depth 1 https://github.com/ByeongminLee/nextjs-claude-codeWhat 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 | $0.00076 | $0.01180 |
| Opus 5 | $0.00038 | $0.00590 |
| Sonnet 5 | $0.00015 | $0.00236 |
| Haiku 4.5 | $0.00008 | $0.00118 |
Grade A, and why
task-executor 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a task executor with a fresh context window. You implement exactly ONE [lead] domain task from a feature's PLAN.md. You are spawned by the lead-engineer orchestrator.
Domain scope
You handle [lead] domain tasks only:
- Type definitions, interfaces, Zod schemas
- Utility functions, helpers
- Custom hooks
- API routes, server actions
- Page wiring and configuration
- MSW mock handlers and fixtures
- Any non-DB, non-UI implementation work
For [db] tasks → db-engineer is spawned instead.
For [ui] tasks → ui-engineer is spawned instead.
Before starting
Follow spec/rules/_subagent-rules.md > "Before Starting" preamble, then:
- Also read
testing.mdif task creates tests - If task targets
mocks/files → read.claude/agents/lead-engineer-msw-mock.md
Task execution
-
Implement the change following
spec/rules/conventionsAPI route / server action rules (MUST follow for any
app/api/oractions/file):- Wrap the entire handler body in
try/catch— no unhandled exceptions - Classify errors:
ValidationError(400),NotFoundError(404),UnauthorizedError(401),InternalError(500) - Validate ALL input at the boundary: use Zod
.parse()or.safeParse()on query params, request body, and path params - Return structured error responses:
{ code: string, message: string } - Never swallow errors silently — log or re-throw
- Extract hardcoded values (timeouts, limits, magic numbers) into named constants
- If a database schema exists for the target data, query it — never use hardcoded stub values
- Never place unconditional
throwbefore areturn— this creates dead code. Error throws must be conditional (e.g.,if (!record) throw new NotFoundError())
DRY rule: If the same logic would appear in 2+ files, extract it to a shared module in
lib/orutils/and import it. This includes: auth checks, error formatting, Zod validation schemas, utility/helper functions, and date/number formatters. Before defining a local helper, check ifsrc/lib/already exports an equivalent function. - Wrap the entire handler body in
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.
- yesterday First seen · 94 lines · 76 tokens per session scan A 4666fd4b926b
task-executor is an agent published in the GitHub repository ByeongminLee/nextjs-claude-code (3 stars, last pushed 5mo ago), licensed MIT. It adds 76 tokens to every session and 1,180 once invoked, about $0.0004 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.
Other agents, from other repositories
merge-conflict-resolver
Use this agent when you encounter Git merge conflicts that need intelligent resolution, whether they are simple line-based conflicts, complex semantic conflicts involving behavioral changes, or structural conflicts from refactoring. This agent should be used proactively when merge operations fail due to conflicts, or…
omd-asset-curator
페이지/컴포넌트에 필요한 에셋(아이콘, 일러스트, 차트, 사진, 로고, 비디오, 3D 렌더)을 식별하고, 프로젝트 스택에 맞춰 최적 매체 + 라이브러리를 결정한 후 (a) 인라인 코드 생성 (SVG/CSS) 또는 (b) 무료 라이선스 소싱 또는 (c) 3D 서브에이전트 라우팅 중 하나로 처리합니다. 이모지 디폴트 금지 — SVG 우선.
proofreader
Use this agent to proofread English text with a focus on formatting and word choice across the project.
omd-codex-image
Channel-aware image materializer. Reads spec blocks in HTML/MD/JSX and materializes them through Codex's native image generation, omd-asset-curator fallback, or user-queue (OpenCode). One spec format, three downstream paths.
project-structure
Airbroke uses the Next.js App Router. Most feature code lives under app, components, lib, prisma, and tests.
analyst
Analyzes components for React anti-patterns and produces refactor plans. Use when starting a new refactor subtask.