Eval Implementer — Development agent for Claude Code
Implementer role of the D365FO agent eval loop.
How to install Eval Implementer
Installs to ~/.claude/agents/dynamics365ninja-d365fo-mcp-server-eval-implementer.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/dynamics365ninja/d365fo-mcp-server/HEAD/.claude/agents/eval-implementer.md -o ~/.claude/agents/dynamics365ninja-d365fo-mcp-server-eval-implementer.md Restart Claude Code, or start a new session, for it to be picked up.
What Eval Implementer does
name: eval-implementer description: Implementer role of the D365FO agent eval loop. Runs ON THE VM (mcp-server in full mode + C# bridge) against the Contoso sandbox model. Takes an eval case id, drives the grounded MCP tool path to implement it, builds, scores against the golden/SysTest oracle, writes a corpus record, and rolls back. Use when asked to "run eval case ", "run the implementer", or execute a case end-to-end on the VM. tools: Bash, Read, Grep, Glob, mcp__d365fo-eval__prepare,
Alternatives in Development
- Browser Stealth Agent — Stealth browser automation agent for targets behind Cloudflare, Akamai, Google, DataDome, or PerimeterX bot de 812 ★
- Part 13: Memory Bridge (Give Coding Agents Your Brain) — When you spawn Codex or Claude Code to build something, they start blind 232 ★
- Prompt Post Impl Review — IMPORTANT: As a trycycle subagent, you have no designated skills 205 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Oneshot Implementer
One-shot implementation agent that runs inside a sandbox container on the server. Reads the sealed requirement
Harness Prover
Running-app verifier for harness eval loops. Drives the live feature (browser, API, or CLI) and returns a bina
Daemon Impl
Daemon implementation agent for Go code. Use when the orchestrator invokes daemon implementation or the user w
Web Impl
Web implementation agent for the TypeScript dashboard (Vite, no framework). Use when the orchestrator invokes
Agentapp Reality Checker
Agent 应用最终验收官。在 eval 循环收敛、安全审查完成后,由 agent-dev-team 剧本在 Phase 7 激活。对照能力画像与 eval 达标情况做整体验收,默认判决是"需要返工",只有压倒性证据才判
Persona Synthesizer
Phase 2 of the grounded-persona-eval method — builds 3-4 personas strictly from Phase 1's sourced evidence. Re
Related Skills
Eval Run
Run an eval case end-to-end on the VM (implement → build → score → record → roll back). VM/full-mode only.
GPT Image
A cross-runtime image skill for Claude Code and GitHub Copilot CLI that keeps the original two-mode design, bu
Claude Agents Python
Seven role-scoped Claude Code subagents for solo Python development. Write permissions are deliberately split