Peek CLAUDE.md

Peek CLAUDE.md is an instructions file for coding agents from Gyangu/Peek. It costs 701 tokens per session, scanned A, original, MIT.

A set of instructions for the Peek project that requires design documents to be checked and updated before behavior, structure, or interaction changes are coded.

In plain words
What is it for?
Reviewing existing design records, resolving conflicts with requested changes, recording the decision, and then implementing the agreed design.
Why use it?
It prevents the code from drifting away from the project’s documented decisions and makes disagreements visible before implementation starts.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/gyangu/peek/claude-md
Clone the repo
git clone --depth 1 https://github.com/Gyangu/Peek

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 Peek CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/gyangu/peek/claude-md.svg)](https://agentmods.dev/instructions/gyangu/peek/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/gyangu/peek/claude-md"><img src="https://agentmods.dev/badge/instructions/gyangu/peek/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 701 This file is loaded in full into every session.
When invoked 701 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.00701 $0.00701
Opus 5 $0.00351 $0.00351
Sonnet 5 $0.00140 $0.00140
Haiku 4.5 $0.00070 $0.00070

Measured 4d ago against content hash 07e9387cb48a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Peek 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 4d 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 · 57 lines

How it starts

The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Peek — working conventions

Every change goes through the documents

Any change that touches behaviour, structure or interaction goes through the documents before it goes into the code. The document stays in the repository and is committed alongside the code, as the decision record for that change.

The order is fixed:

  1. Look for conflicts — before starting, read the existing design documents (docs/PLAN.md and the relevant records under docs/design/) and establish whether what is being asked conflicts with what is already designed.
  2. A conflict is settled first — list the conflict explicitly for the user: what the documents decided, what is being asked now, and where the two fail to line up. Stop and wait for the user to say which side wins. Do not pick one and carry on.
  3. Update the documents once it is settled — revise the design record to the agreed conclusion (edit the old one, or write a new one) so that the documents are the single source of truth again.
  4. Then implement against the documents — the code follows the updated record. If the design turns out not to work during implementation, go back to step 2 and settle it again; do not let the code and the documents quietly diverge.

Design documents:

  • Location: docs/design/
  • Naming: YYYY-MM-DD-<lowercase-hyphenated-topic>.md, for example 2026-08-01-chat-session-management.md
  • Language: English. Documents in this repository are written in English. The reason is that the repository is open source and the source comments were always English — 221 comments point at docs/design/, and the documents have to follow so that readers do not hit a language wall between a citation and its target. This includes design records: as of 2026-08-29 the Chinese ones have been translated, reversing the 2026-08-24 rule that said they would not be. Translating a decision record is allowed; rewriting one is not — claims, dates, measurements and struck-through reversals survive a translation unchanged. Conversation with the user stays in Chinese; this rule governs only what is written into the repository.
  • Cover at least:
    1. What this fixes — the current state, the problem, and the boundary of this change (say explicitly what it does not do)
    2. The plan — what changes, which files and modules it touches, the data structures and how state flows
    3. Trade-offs — the other approaches considered, and why they lost
    4. Verification — how to confirm the change is right (manual steps or tests)

Read the full file on GitHub · 57 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. 4d ago First seen · 57 lines · 701 tokens per session scan A 07e9387cb48a

Subscribe to this mod's changes

Peek CLAUDE.md is an instructions file published in the GitHub repository Gyangu/Peek (0 stars, last pushed 6d ago), licensed MIT. It adds 701 tokens to every session, about $0.0035 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 · 5,182 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

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.

github/spec-kit · 7,104 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

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