Skip to content
KitploitKITPLOIT
工具漏洞利用博客
Log in
提交
工具漏洞利用博客
提交

黑客、渗透测试和网络安全工具,武装您的安全武器库!

Kitploit 是一个黑客、网络安全和渗透测试工具的目录。发现最新的项目更新,查找漏洞、分析系统、自动化测试并加强你的安全。

订阅源联系隐私© 2026 Kitploit

工具目录

分类

查看所有分类
Loading categories
cs440-langgraph-nosql-demo — Local classroom demo of NoSQL operator injection in LangGraph's MongoDBSaver (GHSA-98xf-r82g-9mhx / CVE-2026-48121). Fictional data only. | Kitploit
工具/GitHubGitHub/decker757/cs440-langgraph-nosql-demo
Vulnerability AnalysisCode AnalysisWeb Application ExploitationLearning & EducationAPI SecurityDatabase SecurityLabs & Practice
GitHubdecker757/cs440-langgraph-nosql-demo

cs440-langgraph-nosql-demo

Local classroom demo of NoSQL operator injection in LangGraph's MongoDBSaver (GHSA-98xf-r82g-9mhx / CVE-2026-48121). Fictional data only.

查看仓库
11天前尚未审核
分享
内容在请求的语言中不可用。显示英文版本。

CS440 · NoSQL Injection in LangGraph's MongoDBSaver

A small, local, browser-based demonstration of a real, published security advisory: unvalidated checkpoint identifiers in the LangGraph MongoDBSaver integration let an attacker smuggle a MongoDB query operator into an identifier field and read another user's saved conversation.

  • Advisory: GHSA-98xf-r82g-9mhx / CVE-2026-48121 (https://github.com/langchain-ai/langgraphjs/security/advisories/GHSA-98xf-r82g-9mhx)
  • Related issue: #2351 (https://github.com/langchain-ai/langgraphjs/issues/2351)
  • Vulnerable package: @langchain/[email protected]
  • Patched package: @langchain/[email protected]

Everything runs on your own machine with fictional Alice and Bob data only. No external targets, real accounts, cloud services, or API keys.

This is an integration flaw, not a flaw in MongoDB's query engine. MongoDB is one optional storage backend for LangGraph checkpoints. The bug is that the saver built a query straight from caller-supplied identifier fields without checking they were strings. This is not prompt injection: the attack enters the identifier fields, never the chat message.


What the demo shows

The screen is laid out to be read from the back of a room. For every trial it reveals, left to right:

  1. What Alice sends: the actual submitted identifiers (configurable).
  2. What MongoDB is asked: the actual query, captured live through the driver's command monitoring, with a plain-English reading of it. Or, if the request was rejected, a clear statement that nothing reached MongoDB.
  3. What Alice gets back: the actual restored state, or the actual error.

Below those, In the database shows the two stored chats and which ones the query matched, and Results so far builds up a comparison of every trial.

#TrialVersionthread_id sentOutcome
1Normal request1.3.0 vulnerable"alice-thread"Returns Alice
2Injection attack1.3.0 vulnerable{ "$gt": "" }Returns Bob
3Same attack, patched1.3.1 patched{ "$gt": "" }Rejected before any lookup
4Normal request, patched1.3.1 patched"alice-thread"Returns Alice (fix is not over-broad)
5Bob's real ID, patched1.3.1 patched"bob-thread"Returns Bob (the patch checks types, not ownership)

Trial 5 is not a LangGraph bug. It shows what a type check cannot do: the ID is valid text, so deciding whether Alice may read it is the application's job.


Prerequisites

  • Node.js >= 22 (node --version)
  • Docker Desktop running (for the local MongoDB container)
  • A bash shell for setup.sh: macOS, Linux, or Windows with Git Bash or WSL. The npm run commands also work from PowerShell.

Install dependencies once with npm install.

The two saver versions are installed side by side as npm aliases (see package.json):

"mongodb-saver-vulnerable": "npm:@langchain/[email protected]",
"mongodb-saver-patched":    "npm:@langchain/[email protected]"

Exact commands

./setup.sh          # start: check prereqs, start MongoDB, seed the data, serve the UI
./setup.sh test     # run the five trials as integration tests
./setup.sh logs     # show MongoDB's own log of each trial's query (add -f to follow)
./setup.sh stop     # stop and remove ONLY this demo's MongoDB container + volume

Then open: http://127.0.0.1:4400. You should see "Local server ready" at the top right. Press Ctrl+C to stop the web server.

Without bash, the same steps are npm run db:up, npm start, npm test, and npm run db:down. These use this project's Compose file (compose.yaml, project name cs440-nosql-demo) and touch only this demo's container and volume.


Running the trials

Alice is the simulated requester. The chat message is always the literal text Continue my chat. Only the identifier fields change between trials. Click the numbered buttons in order, 1 to 5.

A trial finishes in milliseconds, so the page reveals the finished result one panel at a time, about a second apart. That pause is display pacing only, not query time.

Reset re-seeds the fictional data (Alice first, Bob last) and clears the results table. Download evidence saves the last run as JSON. Every run is also appended to evidence/runs.jsonl.

Why Bob comes back in Trial 2

{ "$gt": "" } means "any non-empty thread_id", which broadens Alice's lookup to every stored thread. The saver then applies .sort("checkpoint_id", -1) and .limit(1), both visible in the captured query, so it returns the single newest checkpoint. Bob is seeded last, so that checkpoint is Bob's.

Important nuance: a broadened lookup returns the latest matching checkpoint, not every record and not an arbitrarily chosen victim.


Plain-English: thread_id and checkpoint_ns

LangGraph can save the state of a conversation so it can be restored later. Each saved snapshot is a checkpoint. To find the right checkpoint, the saver looks it up by a few identifier fields you pass in config.configurable:

  • thread_id: which conversation. Normally a plain string such as "alice-thread".
  • checkpoint_ns: checkpoint namespace. A sub-label used to separate checkpoints within one conversation. For a simple top-level chat it is "".
  • checkpoint_id (optional): which exact snapshot. If you don't provide one, the saver returns the latest checkpoint for that thread.

Version 1.3.0 dropped these values straight into a MongoDB filter. MongoDB treats a field whose value is an object like { "$gt": "" } as an operator, not as a literal id. Version 1.3.1 rejects any identifier field that is not a string.


How faithful is this? (what actually runs)

  • Real LangGraph. Seeding calls graph.invoke(...) on a compiled StateGraph, and each trial calls graph.invoke(...) again: the real restoration path that internally calls the saver's getTuple.
  • Real, unmodified saver source. The two saver versions are the actual published 1.3.0 and 1.3.1 packages, loaded via npm aliases.
  • Real captured query. The query shown is captured from the Node driver's commandStarted monitoring event.
  • Deterministic node, clearly labelled. The single graph node returns a fixed canned reply and never calls an LLM. This keeps trials reproducible.

Disclosed simplifications:

  • During a trial the saver is wrapped so that writes (put / putWrites) are suppressed while reads (getTuple / list) run the real, unmodified code. This keeps the seeded data identical between runs. Seeding uses the fully unwrapped saver, so the stored checkpoints are written by the genuine saver. See src/saver.js (readOnly) and src/demo.js.
  • The trial buttons choose one of three preset identifier sets (src/fixture.js), which go straight into the real graph.invoke. The demo does not expose a free-form public endpoint. An app that forwards client-supplied IDs without checking them is the affected scenario the advisory describes.
  • The In the database strip comes from separate read-only aggregates, not from the saver. They are not find commands, so they appear in neither the captured query nor the rows of ./setup.sh logs.

Database-side evidence (MongoDB's own log)

The query panel shows the command the app captured. For evidence that does not rely on the app's own word, compose.yaml starts mongod with --slowms 0, so it logs every operation, and this prints those log entries:

./setup.sh logs        # each trial's checkpoint query
./setup.sh logs -f     # follow live

Lookups from seeding return nothing and are counted but not listed.

下载工具

最受欢迎

查看全部 →

发现我们社区最常用的工具。

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 终极 OSINT 工具包:调查图谱、置信度评分、别名家族、地图、时间线和报告。49 个模块、400+ 个站点,无需 API 密钥。

    871
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock 一键执行应用(CVE-2026-43499)

    548
  • 3

    ARTEX

    baarai/artex

    基于LLM的多智能体自主渗透测试平台,具备规划器/执行器智能体、资产与探索图谱、MITM流量记录以及漏洞复测功能。

    383
  • 4

    redforge

    cybersecurity91/redforge

    浏览器端红队工具包,包含 102 个离线工具,用于侦察、载荷生成、凭据攻击、加密、编码和交战报告。

    361
  • 5

    looking-glass

    anonghosty/looking-glass

    CSIRT 平台,可将 Shadowserver 报告导入 MongoDB,并将其重放为交互式地图、分析、发现、公告和高管仪表板。

    305
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    AI原生网络安全的行动系统——在这里,意图转化为受治理的执行,证据沉淀为运营记忆,每一次操作都让下一次更出色。

    280

探索所有工具

浏览我们的工具集合

查看所有工具 →