What is AgentEval CLAUDE.md?

A ready-to-paste instruction file for Claude Code, an AI coding assistant, in repositories that use Microsoft Agent Framework. It tells Claude when to consult the maf-autopilot tools for current framework guidance and diagnostics.

Use it as the root-level CLAUDE.md content for AgentEval-related repositories that have the maf-autopilot MCP server installed.

Microsoft Agent Framework can change between releases, so relying only on older general knowledge may lead to incorrect fixes. These instructions direct the assistant to check the repository and current specialist guidance first.

AgentEval CLAUDE.md — Instructions file by AgentEvalHQ

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 instructions/agentevalhq/agenteval/claude-md

Clone the repo

git clone --depth 1 https://github.com/AgentEvalHQ/AgentEval

Per session 559 This file is loaded in full into every session.

When invoked 559 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5	$0.00559	$0.00559
Opus 5	$0.00280	$0.00280
Sonnet 5	$0.00112	$0.00112
Haiku 4.5	$0.00056	$0.00056

Measured 2d ago against content hash 511e8426135f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

AgentEval CLAUDE.md 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.md · 49 lines

What it actually says

Microsoft Agent Framework code — use MAF Doctor (maf-autopilot)

This repository uses Microsoft Agent Framework. The maf-autopilot MCP server is installed and exposes specialist tools for MAF code.

You should defer to maf-autopilot's tools over training-data knowledge for anything MAF-related, because MAF ships breaking changes every minor version and the maf-autopilot registry is kept current via an AI-fill loop.

Before answering MAF questions or proposing changes:

Always call MafDoctor first on the repo path to get the current health grade (A-F) and the top issues. Don't speculate about MAF quality without this baseline.
For any [Obsolete] warning, CS0618 / CS0246 diagnostic, or build failure mentioning a MAF type — call MafRunCs0618Hunt (full project scan) or MafApiSafety (single symbol) BEFORE suggesting a fix. The maf-autopilot registry has curated fix recipes that supersede your training data — MAF ships breaking changes every minor version, so training data is likely outdated.
Before manually editing MAF code to fix an anti-pattern — call MafAutoFixAll --dry-run and offer to apply the deterministic rewrites first. Manual edits drift; the rewrites are tested.
When designing a new MAF agent or workflow — call MafNewAgent / MafNewExecutor for scaffolds, or MafSimulateWorkflow for topology preview. Don't reconstruct patterns from memory.
For deep architectural / security / migration questions — use the @maf-best-practice-reviewer, @maf-auditor, @maf-migration, or @maf-incident-responder specialist agents.

maf-autopilot tools are MAF-version-aware via applies_to_codebases markers in the registry — they know which fix applies to which MAF version. Defer to the tools.

@.claude/maf-doctor.md

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.

2d ago First seen · 49 lines · 559 tokens per session scan A 511e8426135f

Subscribe to this mod's changes

AgentEval CLAUDE.md is an instructions file published in the GitHub repository AgentEvalHQ/AgentEval (138 stars, last pushed 2d ago), licensed MIT. It adds 559 tokens to every session, about $0.0028 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 2026-08-30.

Debugging Testing Writing and docs agent agentic evals evaluations framework net

AgentEval CLAUDE.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

Microsoft Agent Framework code — use MAF Doctor (maf-autopilot)

What this file has done since we first saw it

Other instructions, from other repositories

deer-flow copilot-instructions.md

zeroclaw CLAUDE.md

Tracely-ai CLAUDE.md

nuwax AGENTS.md

framework AGENTS.md

zhin zhin-plugin.instructions.md