research-closure-harness: Instructions file for Claude Code

CLAUDE.md

research-closure-harness CLAUDE.md is an instructions file for Claude Code from Shiien/research-closure-harness. It costs 600 tokens per session, scanned A, original, MIT.

Repository instructions for a research project that uses a structured process to close open research questions. The project tracks claims, experiments, evidence, and the next permitted investigation in a linked graph.

In plain words
What is it for?
Use them when working on research-closure-harness to inspect project state, find the next allowed probe, record experiment outcomes, and determine when a claim is sufficiently resolved.
Why use it?
They prevent uncontrolled scope changes and unsupported conclusions by requiring a frozen question, written hypotheses, decision records, and progress checks at each session.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md.

This is Shiien/research-closure-harness's own configuration. It tells Claude Code how to work on research-closure-harness itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything research-closure-harness configures →

Reuse

Borrowing it

Nothing to install: this file belongs to Shiien/research-closure-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.

Copy the file
curl -O https://raw.githubusercontent.com/Shiien/research-closure-harness/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/Shiien/research-closure-harness

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.

agentmods badge for research-closure-harness CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/shiien/research-closure-harness/claude-md/github.svg)](https://agentmods.dev/instructions/shiien/research-closure-harness/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/shiien/research-closure-harness/claude-md"><img src="https://agentmods.dev/badge/instructions/shiien/research-closure-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.

agentmods 80×15 button for research-closure-harness CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/shiien/research-closure-harness/claude-md"><img src="https://agentmods.dev/badge/instructions/shiien/research-closure-harness/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 600 This file is loaded in full into every session.
When invoked 600 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00600 $0.00600
Opus 5 $0.00300 $0.00300
Sonnet 5 $0.00120 $0.00120
Haiku 4.5 $0.00060 $0.00060

Measured 9d ago against content hash 47843c908e05, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

research-closure-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 9d 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 · 99 lines

How it starts

The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Research Closure Operating Rules

This repository is managed in graduation-mode research closure.

At the start of every session:

  1. Read .research/state.json.
  2. Read the active sprint and experiment logs in .research/logs/.
  3. Run python tools/research_closure.py guard.
  4. Run python tools/claim_graph.py frontier — the claim graph is the engine, and the ready frontier is what decides which probe may run next.
  5. Run python tools/research_closure.py next to see the event the harness expects.
  6. State the frozen claim, this session's deliverable, and what is out of scope, plus the ready frontier and whether the resolution map is already determined.

For human progress tracking, python tools/research_closure.py dashboard opens the interactive DAG view (theory → variables/edges → probes, outcomes, frontier, resolution map, event log).

Hard workflow constraints

  • One active project question.
  • One frozen sprint claim.
  • One primary active experiment.
  • No experiment without a written hypothesis and kill criterion.
  • No new variant before the previous experiment receives a written decision.
  • No scope change without explicitly revising or closing the sprint.
  • New ideas go to backlog by default.

Before implementing a new experiment, either locate its existing experiment card or create one:

python tools/research_closure.py new-experiment ...

After results exist, close it:

python tools/research_closure.py close-experiment ...

Research reasoning discipline

Always distinguish:

  • project question;
  • current claim;
  • experimental hypothesis;
  • implementation/debugging issue.

Do not upgrade an implementation failure into a new research direction without evidence.

When results are weak, prefer:

  1. verify correctness;
  2. test measurement sensitivity;
  3. narrow the claim;
  4. document the negative result;
  5. terminate if the kill criterion is met.

Do not respond to failure by automatically proposing a new representation objective, exploration bonus, benchmark, or theoretical framework.

Read the full file on GitHub · 99 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.

  1. 9d ago First seen · 99 lines · 600 tokens per session scan A 47843c908e05

Subscribe to this mod's changes

research-closure-harness CLAUDE.md is an instructions file published in the GitHub repository Shiien/research-closure-harness (10 stars, last pushed 16d ago), licensed MIT. It adds 600 tokens to every session, about $0.0030 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 2026-08-31.

Related

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.

vercel/next.js · 7,296 tokens

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.

openai/codex · 5,153 tokens

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).

microsoft/vscode · 6,785 tokens

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).

microsoft/vscode · 5,001 tokens

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.

langchain-ai/langchain · 4,469 tokens

deepseek-harness AGENTS.md

AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.

deepseek-ai/deepseek-harness · 3,735 tokens