page

Corpus Environment

environment memory rag knowledge indexing

Corpus Environment

What It Is

Corpus is the document storage and extraction system for Nexus. When you ingest a document, the extracted content goes here for structured storage and search.

Technical Details

MCP Server: /opt/mcp-servers/corpus/mcp_corpus_server.py
Class: CorpusMCPServer
Redis Ports: 6650 (vault), 6651 (operational)
Redis PREFIX: corp:
Stable ID Format: c_XXXX (4 alphanumeric chars)
Version: 2.1.0

Key Differences from Nexus Docs

Corpus	Nexus Docs
Extracted text content	Actual binary files
Hierarchical storage	Flat file storage
Searchable via Quadfecta	Viewable/downloadable
For AI knowledge retrieval	For human file access
corp: prefix	File paths

Tools

corpus.create - Create document entry
corpus.get - Retrieve by ID (c_XXXX or raw)
corpus.search - Quadfecta search
corpus.list - List documents
corpus.ingest - Ingest PDF with page/chapter chunking
corpus.create_pdf - Generate PDF from markdown
corpus.tree - Show hierarchy
corpus.convert - Convert PDF to markdown (Docling)
corpus.extract - Extract text (PyMuPDF)

Hierarchy

Max depth: 3 levels
Parent-child relationships
CDN URLs link to Nexus Docs files
Track refs link to projects

Multi-tenant

No hardcoded users
user parameter required or uses 'system' fallback
Keys: corp:{user}:{timestamp_id}

🌳 View Tree