Borrowing it
Nothing to install: this file belongs to ucb-bar/merlin. 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/ucb-bar/merlin/main/CLAUDE.mdgit clone --depth 1 https://github.com/ucb-bar/merlinWrote 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/ucb-bar/merlin/claude-md)<a href="https://agentmods.dev/instructions/ucb-bar/merlin/claude-md"><img src="https://agentmods.dev/badge/instructions/ucb-bar/merlin/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/ucb-bar/merlin/claude-md"><img src="https://agentmods.dev/badge/instructions/ucb-bar/merlin/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 | $0.05312 | $0.05312 |
| Opus 5.5 | $0.02125 | $0.02125 |
| Sonnet 5.5 | $0.01062 | $0.01062 |
| Haiku 4.5 | $0.00531 | $0.00531 |
Grade A, and why
merlin 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 15d 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 — 282 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository conventions
This file documents the conventions this repo enforces — the generated-output layout, the test
layout, and the documentation model. Read it (and the directory-local AGENT.md files) before making
changes; the same gates run in pre-commit and CI.
Working-tree note: this repo is frequently developed by multiple coding agents against a single working tree. Commit on the currently checked-out branch and avoid switching branches mid-task; use a separate
git worktreeif you need isolation.
Source ownership
src/merlin/ owns the installed compiler core. Optional phase execution, DSE, mining workflows,
and analysis live under packages/*/src/; their historical imports share the merlin namespace
without duplicate implementations. merlin/python/merlin is only a compatibility symlink.
Experiment definitions start at experiments/catalog.yaml; retained research inputs live under
experiments/reference-data/, not generated out/. See docs/reference/repo_structure.md.
Compiler and OOT ownership rule
All target-specific dialect operations, instruction encodings, device kernels and schedules, hardware layout/resource facts, ABI glue and target execution support belong in the target's OOT MLIR dialect repository. Reusable host code generation, packing, requantization, graph/global optimizations, dispatch, buffer ownership, device compilation orchestration and runtime infrastructure belong in Merlin. An optimization selectable independently of the accelerator belongs in Merlin even when first measured on one target. Split mixed changes at an explicit contract: generic mechanism in Merlin, target facts and implementation in OOT. Promote generic OOT prototypes into Merlin and make the provider delegate; do not maintain duplicate implementations. Preserve explicit numeric policy selection and correctness gates.
Target support providers (backends, target oracles, build support and their tests)
belong in the companion OOT repository. A historical support snapshot may be
retained only with exact companion provenance and explicit evaluation scope;
its presence does not establish a default provider or independent qualification.
Fresh compiler authoring admits only its independently qualified runtime and
public source intake. Protected reference backends, support adapters, schedules
and prebuilt compiler answers remain outside its inputs and tools. Removing a
reference provider must not leave instructions to rediscover it through a legacy
default. Compiler candidates stay in their OOT repositories. See
docs/guides/target_resolution.md.
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.
- 15d ago First seen · 282 lines · 5,312 tokens per session scan A b7f6d1d71bc7
merlin CLAUDE.md is an instructions file published in the GitHub repository ucb-bar/merlin (21 stars, last pushed yesterday), licensed Apache-2.0. It adds 5,312 tokens to every session, about $0.0212 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 2026-10-14.
Other instructions, from other repositories
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).
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.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
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 copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.