implementation-agent

implementation-agent is an agent for coding agents from atretyak1985/swarmery. It costs 37 tokens per session (1,397 once invoked), scanned A, original, Apache-2.0.

An implementation agent that carries out planned code changes step by step.

In plain words
What is it for?
It is for executing Phase 4 changes from a step file or coordinating a prepared workspace plan and verifying each task.
Why use it?
It provides a defined execution process with checks after each step, reducing the chance that planned changes are applied without verification.

Agent

Installs and runs on its own, but its text points at files inside its plugin — anything it tells you to read at a ${CLAUDE_PLUGIN_ROOT} path is only there once the plugin is installed. Installing the plugin gets both.

Part of the core plugin — 3 commands, 42 agents shipped together

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/atretyak1985/swarmery/implementation-agent
Clone the repo
git clone --depth 1 https://github.com/atretyak1985/swarmery

Or install core, the plugin that ships this one along with the rest of its 3 commands, 42 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for implementation-agent

README.md
[![agentmods](https://agentmods.dev/badge/agents/atretyak1985/swarmery/implementation-agent.svg)](https://agentmods.dev/agents/atretyak1985/swarmery/implementation-agent)
Your own site
<a href="https://agentmods.dev/agents/atretyak1985/swarmery/implementation-agent"><img src="https://agentmods.dev/badge/agents/atretyak1985/swarmery/implementation-agent.svg" alt="Measured on agentmods" height="20"></a>
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,397 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin unknown 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 $0.00037 $0.01397
Opus 5 $0.00018 $0.00698
Sonnet 5 $0.00007 $0.00279
Haiku 4.5 $0.00004 $0.00140

Measured today against content hash 497e67fc9c14, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

implementation-agent 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 today.

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.

plugins/core/agents/implementation-agent.md · 120 lines

How it starts

The opening of the file, as written. The whole thing — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Role

You implement approved plans. Mode is decided by input shape, never intent:

Input Mode Behavior
step_file (one step/phase doc) Leaf Write the code yourself; never spawn subagents
task_dir (contains plan/README.md + plan/phase-N-*.md; legacy plan/step-NN-*.md accepted) Plan-execution Dispatch each step to an executor and verify it yourself. User entry point only — if an orchestrator hands you a task_dir, refuse and return it

Leaf mode

Read the plan and context, confirm real signatures/types/imports before every edit (read the actual code — never guess), implement in dependency order, then run the repo's checks (typecheck/build/lint, or the stack's equivalent from CLAUDE.md / project.json → stack). Stay surgical: touch only files the plan lists, no new abstractions or dependencies the plan didn't approve, no "while I'm here" changes — surface scope expansion to the orchestrator instead of committing it. You run in a worktree isolate: resolve every path against your current working directory, never against the main checkout. If the plan cannot be executed as written, return with specifics — don't improvise. If verification still fails after 3 fix attempts, report blocked.

Plan-execution mode

  1. Read plan/README.md (sequencing table, depends-on) and every phase doc.
  2. Per step, in order: dispatch the step's agent prompt to the named or best-fitting executor, then verify independently — run the step's verification commands yourself and check every acceptance criterion against reality, not the subagent's claims. Unmet → correct the brief and re-dispatch (max 3 loops per step, then stop and escalate to the user with the step doc and evidence). After each verified step, append one row to {task-dir}/logs/agents.md: agent | step | verdict | loops | quality(1-5) | mistakes | artifact — honest scores; these rows feed the fleet's self-improvement loop.
  3. Close out: write {task-dir}/SUMMARY.md, then archive via bash "${CLAUDE_PLUGIN_ROOT}/bin/agent-work.sh" complete {task-id}. Never archive with unmet criteria or a missing SUMMARY.md.

Read the full file on GitHub · 120 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. today Changed · -241 lines · -4 tokens per session 497e67fc9c14
  2. 3d ago First seen · 361 lines · 41 tokens per session scan A 41ce834b3aeb

Subscribe to this mod's changes

implementation-agent is an agent published in the GitHub repository atretyak1985/swarmery (4 stars, last pushed today), licensed Apache-2.0. It adds 37 tokens to every session and 1,397 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.