Borrowing it
Nothing to install: this file belongs to benseverndev-oss/goldenmatch. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/benseverndev-oss/goldenmatch/main/CLAUDE.mdgit clone --depth 1 https://github.com/benseverndev-oss/goldenmatchWrote 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.
[](https://agentmods.dev/instructions/benseverndev-oss/goldenmatch/claude-md)<a href="https://agentmods.dev/instructions/benseverndev-oss/goldenmatch/claude-md"><img src="https://agentmods.dev/badge/instructions/benseverndev-oss/goldenmatch/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.
<a href="https://agentmods.dev/instructions/benseverndev-oss/goldenmatch/claude-md"><img src="https://agentmods.dev/badge/instructions/benseverndev-oss/goldenmatch/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.05169 | $0.05169 |
| Opus 5 | $0.02584 | $0.02584 |
| Sonnet 5 | $0.01034 | $0.01034 |
| Haiku 4.5 | $0.00517 | $0.00517 |
Grade A, and why
goldenmatch 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Golden Suite monorepo
Polyglot monorepo: packages/{python,rust,typescript,dbt,actions}. Per-package CLAUDE.md files own package-specific context.
STANDING RULE — flaky test first. The moment a test is observed flaky (intermittent CI red, a re-queue that goes green, a "passes locally / fails in CI" report), STOP the current work and fix the flakiness first, then resume. A flaky test is a real defect — it hides regressions and erodes the merge gate — so it takes priority over the feature in flight, not a "later" backlog item. Fix the ROOT cause (determinism, isolation, resource hardening), not the symptom (no blind retries/rerun bandaids, no loosened assertions); the swallowed-error / non-WAL-SQLite / xdist-shared-state classes are the usual suspects.
North Star (the pull that colors every decision): be the tool any developer reaches for by default for entity resolution. It is never "done" — defaults are re-earned every release. The full statement + the five decision-test commitments (zero-config, scale-invariant correctness, shared-capabilities-conform, approach-the-expert, never-black-box) live in context-network/foundation/project-definition.md. When two changes compete, take the one that advances it.
Architecture governing frame (colors every architectural decision, the way the North Star colors product decisions): GoldenMatch is one product, two engines, many surfaces — an Arrow-native, Rust-authoritative Identity Compute Engine + a transaction-native Identity Control Plane (SQLite/Postgres), exposed through many surfaces governed by spec + conformance. The full frame + roadmap live in context-network/architecture/one-product-two-engines.md (decision context-network/decisions/0047-one-product-two-engines-architecture.md). Read it before any architectural change; the change must conform or amend the frame (doc + 0047) in the same PR. Decision tests every change must pass:
- One authoritative semantic owner per capability — Rust where a clean shared core exists; pure-Python / standalone-TS algorithms are classified, conformance-tested fallbacks, not co-equal sources of truth. Am I adding a second source of truth for behavior that already has an owner?
- Arrow at bulk boundaries, not the universal calling convention — smallest stable primitive for scalar/small calls. Am I paying Arrow/FFI marshaling on a small call where two strings or a
list[str]is cleaner? - Compute vs. control stay distinct — the compute engine is Arrow/batch; the control plane is a transaction-native state machine (stable IDs, merge/split, provenance, audit). Am I forcing stateful identity logic into a columnar kernel, or vice versa?
- Kernelize on measurement — measure the whole path (conversion + orchestration), not a Rust-loop-vs-Python-loop microbench; state the motivation class (perf / semantic single-sourcing / portability / safety / maintainability).
- Conformance defines correctness — a new backend/binding/fallback proves itself against the conformance suite (exact / numerically-equivalent / semantically-equivalent / intentionally-divergent), not by calling the same library. DataFusion/Ray/Sail/Ballista/Spark are replaceable backends, none synonymous with GoldenMatch.
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.
- 5d ago Changed · +1 lines · +90 tokens per session 556f47a8cc27
- 9d ago First seen · 132 lines · 5,079 tokens per session scan A 1824466c2f85
goldenmatch CLAUDE.md is an instructions file published in the GitHub repository benseverndev-oss/goldenmatch (131 stars, last pushed today), licensed MIT. It adds 5,169 tokens to every session, about $0.0258 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.
Other instructions, from other repositories
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.
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.
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).
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).
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.
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.