agentmods.dev

crosscheck: Instructions file for Claude Code

CLAUDE.md

crosscheck CLAUDE.md is an instructions file for Claude Code from hongsungmin0315/crosscheck. It costs 3,459 tokens per session, scanned A, original, MIT.

Development instructions for a system that collects posts from followed Reddit accounts, checks their claims against trusted news sources, and sends a daily summary to Discord. It uses MCP so an AI agent can call the system's tools.

In plain words
What is it for?
Use it when building or maintaining the Reddit-news checking workflow, including claim assessment and Discord digests. TDD is not described here; the instructions focus on evidence-based cross-checking.
Why use it?
It prevents the system from declaring that a claim is true or false when the evidence may be incomplete. Instead, it reports whether sources support, contradict, or do not cover each claim, with links to the evidence.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md; mentions Claude Code.

This is hongsungmin0315/crosscheck's own configuration. It tells Claude Code how to work on crosscheck 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 crosscheck configures →

not rated 0repo 2mo ago A scan 3,459 tokens original MIT
View source ↗ hongsungmin0315/crosscheck
Reuse

Borrowing it

Nothing to install: this file belongs to hongsungmin0315/crosscheck. 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/hongsungmin0315/crosscheck/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/hongsungmin0315/crosscheck

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 crosscheck CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/hongsungmin0315/crosscheck/claude-md/github.svg)](https://agentmods.dev/instructions/hongsungmin0315/crosscheck/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/hongsungmin0315/crosscheck/claude-md"><img src="https://agentmods.dev/badge/instructions/hongsungmin0315/crosscheck/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 crosscheck CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/hongsungmin0315/crosscheck/claude-md"><img src="https://agentmods.dev/badge/instructions/hongsungmin0315/crosscheck/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 3,459 This file is loaded in full into every session.
When invoked 3,459 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 $0.03459 $0.03459
Opus 5 $0.01729 $0.01729
Sonnet 5 $0.00692 $0.00692
Haiku 4.5 $0.00346 $0.00346

Measured 8d ago against content hash 311aa5c1fadd, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

crosscheck 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 8d 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 · 249 lines

How it starts

The opening of the file, as written. The whole thing — 249 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

이 파일은 Claude Code가 작업 시작 시 자동으로 읽는 프로젝트 가이드입니다. 아래의 결정사항과 제약을 항상 지켜주세요.

프로젝트 개요

내가 팔로우하는 Reddit(추후 X/트위터) 계정의 새 글을 수집해서, 그 글이 주장하는 내용을 신뢰할 만한 뉴스 소스와 교차검증(corroboration) 한 뒤, 매일 아침 Discord로 다이제스트를 보내는 자동화 시스템.

핵심 도구는 MCP(Model Context Protocol) 서버로 구현하고, LLM 에이전트가 이 도구들을 호출해서 워크플로를 수행한다.

★ 가장 중요한 설계 원칙 (절대 어기지 말 것)

이 시스템은 글의 진위(진짜/가짜)를 판정하지 않는다. "이건 가짜뉴스다" 같은 단정은 만들지 않는다. 이유:

  • 모델은 진실 여부를 알 수 없고, 틀린 판정은 또 다른 허위정보가 된다.

대신 교차검증 결과만 제시한다. 각 주장(claim)에 대해:

  • 확인됨(corroborated) — 신뢰 가능한 소스에서 같은 내용 보도
  • 반박됨(contradicted) — 소스가 반대되는 내용을 보도
  • 미보도(uncovered) — 관련 보도를 찾지 못함

그리고 항상 근거 링크를 함께 제시한다. 최종 판단은 사용자가 한다. 다이제스트의 어조는 단정이 아니라 "참고 정보 묶음"이어야 한다.

이 원칙을 코드로 강제하는 법 (보강)

assess_claim의 LLM 프롬프트에 다음을 못박는다:

  • "제공된 기사 본문/요약에 있는 내용'만' 근거로 판단하라. 네 사전지식·추론으로 사실을 만들지 마라."
  • 출력에 반드시 source_idx(근거가 된 기사 번호)를 포함시키고, 근거가 없으면 uncovered로 보낸다. 근거 없는 corroborated/contradicted는 금지.
  • verdict는 위 3종 + insufficient(판단 불충분)만 허용. 그 외 자유서술 금지.

기술 스택

  • 언어: Python 3.11+
  • MCP: 공식 Python SDK 내장 FastMCP 사용.
    • import: from mcp.server.fastmcp import FastMCP
    • 설치: uv add "mcp[cli]" (또는 pip install "mcp[cli]")
    • ⚠️ 버전 핀: 공식 SDK는 v2가 곧 정식 전환되며 호환성 변경이 예상됨. 당분간 mcp>=1.27,<2 로 상한을 걸어 안정 버전에 고정할 것.
    • ⚠️ 이름 혼동 주의: 별도 PyPI 패키지인 fastmcp(3.x, from fastmcp import ...) 와 다름. 이 프로젝트는 공식 SDK 내장본(mcp.server.fastmcp)을 쓴다. 둘을 섞지 말 것.
    • 정확한 데코레이터/반환 형식은 작업 시점 SDK 문서로 재확인. 일반 형태: @mcp.tool() 데코레이터 + 타입힌트 + 한국어 docstring(도구 설명으로 쓰임).
    • ⚠️ 반환값 함정: 도구가 raw dict를 반환하면 일부 클라이언트(Claude Desktop) 에서 출력이 잘릴 수 있음. 도구 반환은 **문자열(JSON 직렬화 문자열 권장)**로 감싸고, Inspector(mcp dev)와 Claude Desktop 양쪽에서 확인할 것.
  • Reddit 수집: PRAW (개인용 무료 OAuth)
  • 뉴스 검색: search_news 도구 내부에서 처리. 소스는 교체 가능하게 추상화.
    • 시작 후보: Currents API(무료 일 ~600요청, 상업 허용) 또는 NewsAPI 무료 티어. NewsAPI 무료는 기사 24시간 지연·비상업 제한이 있으나, 매일 아침 다이제스트 용도엔 지연이 큰 문제가 아님(서버사이드 호출이라 CORS 제한도 무관).
    • 최종 목표(미국주식)로 확장 시 Finnhub/Marketaux 같은 금융 특화 API로 교체.
    • 신뢰 소스 allowlist(아래 "교차검증 품질" 참고)로 필터링.
  • Discord 전송: 봇 없이 Incoming Webhook URL로 POST (httpx 또는 requests)
  • 스케줄링: 로컬 cron / macOS launchd, 또는 schedule 라이브러리
  • LLM 호출: Anthropic SDK (anthropic). 모델은 비용/품질 균형상 Sonnet 기본.
  • 비밀값: 모두 .env로 관리 (.env.example 함께 제공, .env는 .gitignore)

Read the full file on GitHub · 249 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. 8d ago First seen · 249 lines · 3,459 tokens per session scan A 311aa5c1fadd

Subscribe to this mod's changes

crosscheck CLAUDE.md is an instructions file published in the GitHub repository hongsungmin0315/crosscheck (0 stars, last pushed 2mo ago), licensed MIT. It adds 3,459 tokens to every session, about $0.0173 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.

Research
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