Borrowing it
Nothing to install: this file belongs to haruhisa-enomoto/math-research-harness. 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/haruhisa-enomoto/math-research-harness/main/CLAUDE.mdgit clone --depth 1 https://github.com/haruhisa-enomoto/math-research-harnessWrote 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/haruhisa-enomoto/math-research-harness/claude-md)<a href="https://agentmods.dev/instructions/haruhisa-enomoto/math-research-harness/claude-md"><img src="https://agentmods.dev/badge/instructions/haruhisa-enomoto/math-research-harness/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/haruhisa-enomoto/math-research-harness/claude-md"><img src="https://agentmods.dev/badge/instructions/haruhisa-enomoto/math-research-harness/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 | $0.00375 | $0.00375 |
| Opus 5.5 | $0.00150 | $0.00150 |
| Sonnet 5.5 | $0.00075 | $0.00075 |
| Haiku 4.5 | $0.00038 | $0.00038 |
Grade A, and why
math-research-harness 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 13d 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.
What it actually says
CLAUDE.md
This repository's session contract lives in AGENTS.md—read it first, then
follow its task-sensitive boot sequence. Mathematical work reads STEERING.md
and the relevant thread ledger; docs/PROJECT-FOUNDATIONS.md and
threads/INDEX.md are conditional. There is one contract for all agents; this
file is only a Claude-specific entry point.
Memory policy
Do not use the file-based memory system in this repository. Do not read from or
write to the auto-memory directory, and do not create, update, or rely on
MEMORY.md or any individual memory files. Treat memory as disabled.
All durable context lives in the repository itself: AGENTS.md, STEERING.md,
thread ledgers, conditional shared references such as
docs/PROJECT-FOUNDATIONS.md, and the threads/, results/, and data/
trees. If something is worth preserving, write it to the
appropriate repository record rather than to memory.
Rules that are easy to forget
- Use plain, short English with the owner. Avoid long or ornate wording.
- Take every date and note timestamp from the operating system immediately before finalizing it.
- Before any search or batch computation, reread "Search discipline" in
docs/COMPUTATIONS.md: close finite strata with exact counts, reduce by symmetry first, start at the edge of what is already settled, time a pilot before a long job, and do not substitute open-ended sampling for a stratum. - Run long jobs so the harness tracks them (a normal or background tool call),
never detached with
nohup ... &,disown, orsetsid; a detached job sends no completion notice.
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.
- 13d ago Changed · +13 lines · +160 tokens per session 2358183d845e
- 2mo ago First seen · 20 lines · 215 tokens per session scan A 0d67aada0db8
math-research-harness CLAUDE.md is an instructions file published in the GitHub repository haruhisa-enomoto/math-research-harness (16 stars, last pushed 14d ago), licensed MIT. It adds 375 tokens to every session, about $0.0015 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 2026-08-30.
math-research-harness CLAUDE.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| math-research-harness CLAUDE.mdharuhisa-enomoto/math-research-harness | Instructions file | 16 | 14d ago | MIT | A | 375 | 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 |