What is better-rlm CLAUDE.md?

A repository instruction file for Claude Code, covering a recursive-language-model server and the project’s development rules. A recursive language model processes very large inputs in smaller steps.

Following trunk-based development, running the project’s test command, editing the included engine, and starting the server correctly.

It tells the coding agent how this repository is structured, how to test it, and where its vendored engine code lives.

How logunovFGP/better-rlm configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to logunovFGP/better-rlm. 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/logunovFGP/better-rlm/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/logunovFGP/better-rlm

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/logunovfgp/better-rlm/claude-md/github.svg)](https://agentmods.dev/instructions/logunovfgp/better-rlm/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/logunovfgp/better-rlm/claude-md"><img src="https://agentmods.dev/badge/instructions/logunovfgp/better-rlm/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/logunovfgp/better-rlm/claude-md"><img src="https://agentmods.dev/badge/instructions/logunovfgp/better-rlm/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 8,005 This file is loaded in full into every session.

When invoked 8,005 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	$0.08005	$0.08005
Opus 5.5	$0.03202	$0.03202
Sonnet 5.5	$0.01601	$0.01601
Haiku 4.5	$0.00800	$0.00800

Measured 26d ago against content hash 4153cfaa9e90, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

better-rlm 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 26d 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 · 559 lines

How it starts

The opening of the file, as written. The whole thing — 559 lines — stays where its author put it; the contents beside it link to each section on GitHub.

rlm-mcp

MCP server exposing Recursive Language Models over oversized contexts. Fork of eesb99/rlm-mcp. See README.md for the tool surface and docs/ for the architecture notes.

Branching

This repo uses trunk-based development. Read TRUNK-BASED-PATTERNS.md before any code change. Config: trunk-based.json (repo root). Worktrees and long-lived branches are blocked by a PreToolUse guard.

Verify command: uv run --extra dev pytest -q — python is not on PATH here, so it must go through uv, and --extra dev is what pulls in pytest (it lives in the dev extra, so a plain uv run resolves the project without it). Same command the pre-push hook runs: scripts/githooks/pre-push, wired up by the installers via core.hooksPath.

Patching the vendored engine

The engine is vendored source at ./rlm, not a dependency — copied from alexzhang13/rlm at v0.1.3, provenance and the upstream-merge recipe in rlm/UPSTREAM.md. Engine fixes are ordinary edits there; import rlm resolves to this copy (rlms is no longer installed).

One import-time patch remains, and it is not a workaround: better_rlm/auth.py:patch_engine rebinds the engine's AnthropicClient so completions route through our transport, throttle and retry. That is dependency injection — folding it into rlm/ would make the engine import from better_rlm/, inverting the dependency and breaking the engine's standalone use. It stays. It is idempotent, guarded by an _rlmmcp_patched flag.

The Docker REPL exec-protocol fixes that used to live in src/sandbox_patch.py are now the engine's own code (rlm/environments/docker_repl.py): the hardened result marker, the capped locals echo, the atomic state.dill write, the state-load warning, UTF-8 host writes, and a real timeout_s. tests/test_sandbox.py:: test_template_still_carries_every_hardening is the regression guard that replaced the old fail-loudly string match — an upstream merge that reverts the template fails there. What is left of that module is better_rlm/sandbox_reap.py, host-side housekeeping that sweeps sandboxes abandoned by a dead server.

Read the full file on GitHub · 559 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.

26d ago Changed · +32 lines · +513 tokens per session 4153cfaa9e90
1mo ago Changed · +266 lines · +3,960 tokens per session d3cc58447a92
1mo ago Changed · +70 lines · +988 tokens per session a176c80bbcf7
1mo ago Changed · +143 lines · +1,918 tokens per session 6363c1ca53ca
2mo ago First seen · 48 lines · 626 tokens per session scan A a34ae3dbef25

Subscribe to this mod's changes

better-rlm CLAUDE.md is an instructions file published in the GitHub repository logunovFGP/better-rlm (0 stars, last pushed 25d ago), licensed MIT. It adds 8,005 tokens to every session, about $0.0320 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 2026-08-31.

better-rlm CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
better-rlm CLAUDE.mdlogunovFGP/better-rlm	Instructions file	0	25d ago	MIT	A	8,005	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

better-rlm: Instructions file for Claude Code