web-agent-comparison: Instructions file for Claude Code

CLAUDE.md

web-agent-comparison CLAUDE.md is an instructions file for Claude Code from pleasedodisturb/web-agent-comparison. It costs 1,682 tokens per session, scanned A, original, MIT.

A project instruction file for Web-Agent MCP Comparison, a public benchmark that compares browser-automation tools used by Claude Code. It documents the project’s purpose, technical setup, rules, and working conventions.

In plain words
What is it for?
Use it when developing, testing, or documenting the benchmark’s Python scoring, Markdown results, shell orchestration, fixtures, and security constraints.
Why use it?
It gives contributors a shared reference for producing repeatable comparisons instead of relying on undocumented choices. Reproducibility means another person can run the same tests and check the results.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

This is pleasedodisturb/web-agent-comparison's own configuration. It tells Claude Code how to work on web-agent-comparison itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything web-agent-comparison configures →

Reuse

Borrowing it

Nothing to install: this file belongs to pleasedodisturb/web-agent-comparison. 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/pleasedodisturb/web-agent-comparison/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/pleasedodisturb/web-agent-comparison

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.

agentmods badge for web-agent-comparison CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/pleasedodisturb/web-agent-comparison/claude-md.svg)](https://agentmods.dev/instructions/pleasedodisturb/web-agent-comparison/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/pleasedodisturb/web-agent-comparison/claude-md"><img src="https://agentmods.dev/badge/instructions/pleasedodisturb/web-agent-comparison/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,682 This file is loaded in full into every session.
When invoked 1,682 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.01682 $0.01682
Opus 5 $0.00841 $0.00841
Sonnet 5 $0.00336 $0.00336
Haiku 4.5 $0.00168 $0.00168

Measured 7d ago against content hash 64ffc78d2432, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

web-agent-comparison 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 7d 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 · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Project

Web-Agent MCP Comparison

A public benchmark of browser-automation MCP servers driven by Claude Code. Stage 1 of a 3-stage pipeline that ends in production agent tooling: this repo scores candidate MCPs on standardized job-application fixtures, the winners graduate into the private terminal-craft toolkit (Stage 2), which is then wired into the Kestrel and Eyas job-hunting agents (Stage 3). Reproducible methodology so external readers can clone, run, and confirm the scores.

Core Value: Pick the right browser MCP(s) for production agent use, backed by reproducible scores on the same fixtures every candidate is measured against. If everything else fails, the comparison matrix and the graduate-to-toolkit recommendation are what must exist at the end.

This file does not restate global rules — read ~/.claude/CLAUDE.md first.

Constraints

  • Tech stack: Python 3 (scoring), Markdown (results), shell (test orchestration). No framework — keep it dogfood-friendly.
  • Reproducibility: Methodology must be runnable by a third party with only the public repo. No internal-only fixtures, no rbw-gated secrets in the core flow.
  • API keys: FIRECRAWL_API_KEY required for firecrawl MCP (in rbw under firecrawl.devFirecrawl_API). If absent, partial scoring (6/7) is acceptable per G-703 AC. No other paid keys. check_prereqs.sh reads the var straight from the environment — export it before make bench (e.g. export FIRECRAWL_API_KEY=$(rbw get firecrawl.dev --field Firecrawl_API)).
  • Sandbox-only MCPs: cloakbrowser is closed-source binary touching cookies — never point at authenticated host pages. Tested only against the public Greenhouse + Ashby fixtures.
  • Public repo: .mcp.json is committed and visible. Acceptable for a research repo; the candidate list IS the research artifact.
  • Linear traceability: G-703 is the umbrella ticket (estimate=16 = break-before-cycle signal). Splits into ~7 per-MCP scoring tickets + 1 synthesis ticket before pulling into a cycle.
  • Cross-machine: Mac Mini has all 7 binaries installed; MacBook parity not yet verified. The .mcp.json will silently fail to spawn missing binaries.

Read the full file on GitHub · 83 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. 7d ago First seen · 83 lines · 1,682 tokens per session scan A 64ffc78d2432

Subscribe to this mod's changes

web-agent-comparison CLAUDE.md is an instructions file published in the GitHub repository pleasedodisturb/web-agent-comparison (2 stars, last pushed 5d ago), licensed MIT. It adds 1,682 tokens to every session, about $0.0084 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.

Related

Other instructions, from other repositories

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,469 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens