What is ade-bench CLAUDE.md?

Repository instructions for ADE-Bench, a framework that evaluates AI agents on data-analysis and data-engineering tasks. It uses dbt, SQL, databases, containers, and result-checking queries rather than general software tests alone.

Use them when adding or changing benchmark tasks, datasets, Docker environments, dbt workflows, SQL result checks, or experiment runs.

They explain the project’s benchmark structure and how database-backed tasks are executed and validated.

How dbt-labs/ade-bench configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to dbt-labs/ade-bench. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/dbt-labs/ade-bench/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/dbt-labs/ade-bench

Made for: Claude Code.

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/dbt-labs/ade-bench/claude-md/github.svg)](https://agentmods.dev/instructions/dbt-labs/ade-bench/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/dbt-labs/ade-bench/claude-md"><img src="https://agentmods.dev/badge/instructions/dbt-labs/ade-bench/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/dbt-labs/ade-bench/claude-md"><img src="https://agentmods.dev/badge/instructions/dbt-labs/ade-bench/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 1,359 This file is loaded in full into every session.

When invoked 1,359 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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 5.1	$0.01359	$0.01359
Opus 5.5	$0.00544	$0.00544
Sonnet 5.5	$0.00272	$0.00272
Haiku 4.5	$0.00136	$0.00136

Measured 2mo ago against content hash 50eae97107c9, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

ade-bench 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 2mo 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 · 184 lines

How it starts

The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md - ADE-Bench Development Notes

Project Overview

ADE-Bench (Analytics and Data Engineering Benchmark) is a benchmarking framework for evaluating AI agents on data analyst tasks. It's modeled after terminal-bench but specialized for dbt and SQL workflows.

Key Differences from Terminal-Bench

Containers: Include dbt + database (DuckDB/SQLite/PostgreSQL) instead of general dev environments
Tests: SQL queries validate results instead of pytest
Tasks: Focus on data transformations, aggregations, and analytics

Architecture

Core Components

harness.py: Main orchestrator for running benchmarks
trial_handler.py: Manages individual task execution
sql_parser.py: Validates task results using SQL queries
docker_compose_manager.py: Handles dbt/database containers

Directory Structure

ade-bench/
├── ade_bench/          # Core Python package
├── tasks/              # Individual task definitions
├── docker/base/        # Base Docker images
├── shared/defaults/    # Default configurations
├── experiments/        # Benchmark results
└── datasets/           # Dataset configurations

Task Structure

Each task contains:

task.yaml: Metadata and configuration
dbt_project/: dbt project files
tests/: SQL validation queries
expected/: Expected query results
solution.sh: Reference solution
Dockerfile: Container setup (optional, uses defaults)

Running Commands

Create a new task:

uv run wizard

Run benchmarks:

# With sage agent
uv run scripts_python/run_harness.py --agent sage --task-ids task1 task2

Key Parameters:

--agent: Agent type (sage, claude, codex, gemini, etc.)
--model: LLM model for AI agents
--dataset-config: YAML file defining task collection
--n-concurrent-trials: Parallel execution (default: 4)
--no-rebuild: Skip Docker rebuilds
--cleanup: Remove Docker resources after run
--plugin-set: Plugin set names from experiment_sets/plugin-sets.yaml (space-separated). Controls skills, MCP servers, and allowed tools. Defaults to none (no plugins).

Read the full file on GitHub · 184 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.

2mo ago First seen · 184 lines · 1,359 tokens per session scan A 50eae97107c9

Subscribe to this mod's changes

ade-bench CLAUDE.md is an instructions file published in the GitHub repository dbt-labs/ade-bench (125 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 1,359 tokens to every session, about $0.0054 per session on Opus 5.5. 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.

ade-bench CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
ade-bench CLAUDE.mddbt-labs/ade-bench	Instructions file	125	1mo ago	Apache-2.0	A	1,359	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,516	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
vscode model-prompts.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	2,022	GitHub Copilot

ade-bench: Instructions file for Claude Code