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.
npx agentmods add instructions/bence-toth/agent-benchmark/agents-mdgit clone --depth 1 https://github.com/bence-toth/agent-benchmarkWhat 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 | $0.00989 | $0.00989 |
| Opus 5 | $0.00495 | $0.00495 |
| Sonnet 5 | $0.00198 | $0.00198 |
| Haiku 4.5 | $0.00099 | $0.00099 |
Grade A, and why
agent-benchmark AGENTS.md scanned grade A with 1 finding 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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- **Minimal dependencies.** Only `js-yaml` is a runtime dependency. Everything else uses Node.js built-ins (`fs/promises`, `child_process`, `path`, `readline`, `node:test`). How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent instructions for agent-benchmark
Project overview
agent-benchmark is a Node.js CLI tool (ESM, no build step) that runs a coding task across multiple git worktrees – each with different AI assistant config files – and produces a comparison report. It is published to npm as agent-benchmark.
The entry point is bin/cli.js. All logic lives in lib/. Tests are flat in test/ (no subdirectories).
Architecture
bin/cli.js – subcommand router
lib/
args.js – flag parsing (parseInitArgs, parseRunArgs, parseReviewArgs, parseCopilotReviewArgs)
config.js – YAML load + validate (loadConfig)
copilot-reviewer.js – GitHub Copilot review orchestration (runCopilotReview)
init.js – agent-benchmark init (initBenchmark)
metrics.js – stream-json parsing (parseMetrics, createCollector)
pricing.js – cost normalization accounting for cache pricing (normalizedCost)
report.js – table render + file write (generateReport, writeResultFiles, listResults, showResult, resolveTimestamp, loadResultSet)
review-prompt.js – review prompt construction (buildReviewPrompt)
review-report.js – review report render + file write (printReviewReport, writeReviewFiles)
reviewer.js – Claude-based review orchestration (runReview)
runner.js – Claude process orchestration (runBenchmark)
worktree.js – git worktree lifecycle (worktreePath, branchName, createWorktree, applyConfigOverlay, commitChanges, discardChanges, removeWorktree, pushBranch, getBaseCommit, getDiffStats)
commands/
copilot-review.js – copilot-review subcommand
init.js – init subcommand
results.js – results subcommand
review.js – review subcommand
run.js – run subcommand
Key constraints
- ESM only. All files use
import/export. Norequire(). No.mjsextensions needed because"type": "module"is set inpackage.json. - No build step. Source runs directly. Do not add transpilation.
- Minimal dependencies. Only
js-yamlis a runtime dependency. Everything else uses Node.js built-ins (fs/promises,child_process,path,readline,node:test). - Node 18+ only. Use the built-in test runner (
node:test),fs/promises,Array.findLast, and other Node 18 APIs freely.
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 · 79 lines · 989 tokens per session scan A 6b41f48c16cc
agent-benchmark AGENTS.md is an instructions file published in the GitHub repository bence-toth/agent-benchmark (2 stars, last pushed 3mo ago), licensed MIT. It adds 989 tokens to every session, about $0.0049 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
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.
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).
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.
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.
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).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.