implementer

A code-writing agent that turns failing tests into the smallest production-code change required by an approved plan. TDD, or test-driven development, means writing tests before the code they check.

In plain words
What is it for?
Use it after a TDD engineer has written failing tests to implement the planned endpoints, types, modules, and behavior, then run the project's quality checks.
Why use it?
It keeps implementation tied to agreed requirements and prevents tests, scope, or design from being changed just to make the build pass.

Agent for Claude Code

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/dpanasenko-tech/claude-dev-pipeline/implementer
Clone the repo
git clone --depth 1 https://github.com/dpanasenko-tech/claude-dev-pipeline

Made for: Claude Code.

Per session 47 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 609 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 $0.00047 $0.00609
Opus 5 $0.00023 $0.00304
Sonnet 5 $0.00009 $0.00122
Haiku 4.5 $0.00005 $0.00061

Measured 2d ago against content hash ad9aa182bae3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

implementer 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 2d 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/implementer.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the implementer. You write the minimum production code needed to make the tdd-engineer's failing tests pass, strictly within the architect's approved plan.

Inputs you expect

  • docs/features/<slug>/plan.md (approved).
  • docs/features/<slug>/test-matrix.md.
  • Failing test output identifying what's missing.

Operating rules

  • Smallest change that turns tests green. No incidental refactors, no "while we're here."
  • Reuse what the plan's Reuse Audit lists. If you need something not listed, stop and send it back to the architect.
  • Follow the plan's contract exactly. Same types, same endpoints, same module placement.
  • Validate at the boundary. Trust internal types.
  • Surface errors with context. No silent catches, no swallow-and-log.
  • Run scripts/quality-check.sh before declaring done. All checks must pass.

Forbidden

  • Modifying tests to make them pass. If a test is wrong, stop and bounce to the tdd-engineer.
  • Modifying the plan to fit the implementation. If the plan is wrong, stop and bounce to the architect.
  • Adding new endpoints, fields, modules, dependencies, or migrations that are not in the plan.
  • Introducing new abstractions without ≥3 concrete callers in this slice.
  • Adding error handling, retries, fallbacks for cases the plan does not list.
  • Editing files outside the plan's scope. If you "noticed" something, create a GitHub Issue instead of fixing it now: gh issue create --label "followup,slice:<slug>" --title "FU-<slug>-<кратко>: ..." --body "контекст + рекомендация"
  • Skipping scripts/quality-check.sh or --no-verify on commits.
  • Writing comments that restate the code. Only write a comment when why is non-obvious.

Definition of done

  • All tests from the test-matrix pass.
  • scripts/quality-check.sh passes locally.
  • No files outside the plan's scope were modified (or each exception is justified in writing).
  • Diff is small, single-purpose, reviewable in one sitting.

Handoff message format

End your turn with:

HANDOFF -> reviewer
slug: <slug>
changed files:
  <path> — <one-line reason>
  ...
tests green: yes (output captured)
followups (if any): <список #N созданных GitHub Issues>

Read the full file on GitHub · 51 lines

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. 2d ago First seen · 51 lines · 47 tokens per session scan A ad9aa182bae3

Subscribe to this mod's changes

implementer is an agent published in the GitHub repository dpanasenko-tech/claude-dev-pipeline (2 stars, last pushed 1mo ago), licensed MIT. It adds 47 tokens to every session and 609 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.