Skip to main content
subagent CX Codex

implementer · mralexivy — Claude Code subagent: setup and use

implementer is a Claude Code subagent that carries out one workstream from an approved plan. It edits its assigned files, adds tests, and runs validation.

MIT ★ 80 updated 2026-11-17

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/agents/implementer.md
Open on GitHub

What implementer does

implementer handles one workstream from an approved plan. It edits the files assigned to that workstream, writes or updates tests, then runs the build and validation commands. Its tools are Read, Edit, Write, Grep, Glob, and Bash.

It follows the plan rather than redesigning the architecture. If the plan leaves a contract or design question unresolved, the subagent stops and reports one precise question with file-and-line evidence. After two attempts to fix a validation failure, it reports what still fails.

Changes stay in the working tree. The subagent avoids destructive commands such as git reset --hard and rm -rf. It installs dependencies only when the plan calls for them.

What it helps with

  • implement one workstream from an approved plan
  • add tests for newly implemented behavior
  • run the build and validation commands for a workstream
  • escalate unresolved contract questions with file-and-line evidence

Example prompts for the agent

  • Implement the workstream that handles error reporting from the approved plan.
  • Add tests for the planned behavior and run its validation commands.
  • Stop and ask a precise question if the plan leaves a contract unresolved.

How to install implementer

Manually: put the files in place

.claude/agents/implementer.md

For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean

Questions about implementer

How do I install implementer in Claude Code?

Add the subagent file at .claude/agents/implementer.md.

What does implementer do?

It carries out one workstream from an approved plan, edits its assigned files, writes tests, and runs the build and validation commands.

When does implementer escalate instead of continuing?

It stops when it encounters an incompatible contract, hidden coupling, a security concern, or an unanswered design question. It reports a precise question with file-and-line evidence.

Source text

implementer.md · MIT
---
name: implementer
description: AutoMode implementer (Sonnet). Implements one workstream of an approved AutoMode plan, writes and updates tests, and runs the build and tests for it. Edits only the files the workstream owns.
tools: Read, Edit, Write, Grep, Glob, Bash
model: sonnet
maxTurns: 60
color: green
---

You are a AutoMode implementer. You carry out one workstream of an approved plan.

Rules:
- Follow the plan. If a small step is wrong, do the closest correct thing and list it under `deviations`. Do not redesign the architecture.
- Escalate instead of guessing. If you hit an incompatible contract, hidden coupling, a security concern, or a design question the plan does not answer, stop, set `escalation.needed` to true, and give one precise question with `path:line` evidence. The architect answers only that question.
- Keep `notes` factual. The reviewer does not see them; it reads the diff itself.
- Edit only the files your workstream owns. If you must touch another file, keep the change minimal and list it in `files_changed`.
- Match the style, naming, and patterns of the code around you.
- Write or update tests for the behavior you add.
- Run the validation commands for your workstream. If a command fails because of your change, fix it. Stop after two fix attempts and report what still fails.
- Never run `git commit`, `git push`, `git reset --hard`, `git checkout -- <path>`, `git clean`, `git stash drop`, `rm -rf`, or other destructive commands. Leave every change in the working tree.
- Do not install new dependencies unless the plan says so.
- Return a short summary. Do not paste diffs.

© mralexivy · MIT · mralexivy/AutoMode

Similar subagents

All subagents for Agent setup: memory, context, orchestration →

implementer by other authors comes first.

implementer Jan De Dobbeleer The implementer subagent for Claude Code implements one agreed task from a triaged waku issue. The orchestrator can run it alongside other agents when their tasks are independent. CX Codex ★ 115repo stars gem-implementer official github gem-implementer is a GitHub Copilot subagent for implementing features, fixing bugs, and refactoring code. It does not review its own changes. CX Codex +1 ★ 41,430repo stars codex-implementer Sateezg codex-implementer delegates large, clearly scoped coding changes to Codex, then reviews the diff and verifies the result. Useful for repetitive edits across files and bulk generation. CX Codex ★ 195repo stars code-testing-implementer managedcode The code-testing-implementer subagent implements one test-plan phase, writes the test files, and checks whether they build and pass. Use it to carry out a phase or run a build-test-fix cycle. CX Codex ★ 521repo stars code-implementer galdawave code-implementer executes technical tickets through strict atomic phases—research, planning, review, implementation, and refactoring—halting immediately after each step. CX Codex +1 ★ 452repo stars cleanup-implementer alfredoperez cleanup-implementer executes a single repository cleanup work packet. It edits assigned files, runs target checks, and never touches git write operations. CX Codex ★ 110repo stars sales-methodology-implementer OneWave-AI sales-methodology-implementer turns a chosen sales methodology into discovery questions, deal scoring and a rep training plan. Useful when everyone qualifies deals differently. CX Codex +1 382installs

From the same set10

The whole set →
diagnoser diagnoser is a Claude Code subagent that examines output from a failed test, build, or lint run. It identifies a likely cause and returns evidence pointers with a fix plan. CX Codex ★ 80repo stars explorer The explorer subagent for Claude Code gathers repository facts for a planner. It identifies dependencies, build commands, and tests before task planning begins. CX Codex ★ 80repo stars builder builder is a Claude Code subagent that takes a standalone task brief, edits the working tree, and loops through project checks until they pass. Runs for rescue attempts and dedicated build tasks. CX Codex ★ 80repo stars verifier verifier is a Claude Code subagent that runs tests, type checks, and builds after code changes. It reviews the diff for regressions and repairs defects caused by the current work. CX Codex ★ 80repo stars reviewer The reviewer subagent for Claude Code performs an independent audit of code changes. It evaluates completed work against original task requirements and acceptance criteria, identifying critical issues. CX Codex ★ 80repo stars architect architect is a Claude Code subagent that turns discovery summaries into an architecture plan. It assigns file ownership and defines acceptance checks. CX Codex ★ 80repo stars growth growth is a Claude Code skill for AutoMode’s weekly promotion cycle. Ask it to draft social posts, check catalogue submissions, find listings, or report traffic and stars. CX Codex +1 ★ 80repo stars publish-results publish-results is a Claude Code skill that propagates benchmark figures across project docs and site files. Runs before merging release PRs or after altering public metrics. CX Codex +1 ★ 80repo stars

Project sponsors

Companies that keep this analytics open