Skip to main content
Nia has two ways to browse the product:
  • Capabilities: what Nia does for you
  • Source types: what kind of knowledge you want to work with
If you already know you need PDFs, code, documentation, datasets, Slack, or Google Drive, start at Source Types. This page is the capability view.

Source Types

Browse Nia by repositories, docs, PDFs, datasets, spreadsheets, Slack, Google Drive, and local sources.

API Reference

Need the full endpoint and schema reference? Jump straight to the API docs.

​
Capability Overview

Index & Subscribe

Bring knowledge into Nia from repositories, docs, papers, datasets, spreadsheets, Google Drive, and local folders. Use index, manage_resource, and auto_subscribe_dependencies.

Search, Read & Explore

Search semantically, grep precisely, read files and sections, and inspect structures. Use search, nia_read, nia_grep, nia_explore, and nia_package_search_hybrid.

Document Agent

Deploy an autonomous AI agent into any indexed PDF or document. It researches using tools (search, read, navigate), delivers cited answers, and supports structured output via JSON schemas. Start with Document Agent.

Data Extraction

Extract structured data from PDFs into JSON — tables, financial records, engineering specs. Define a schema and get clean records. Start with Data Extraction.

Research & Analysis

Ask Nia to compare options, investigate a problem, or review your code against documentation. Use nia_research, nia_advisor, Tracer, and Oracle Research Agent.

Search Without Indexing

Need quick answers from public code right now? Use Tracer for GitHub repositories and nia_package_search_hybrid for package source code.

Local Sync

Keep local folders, databases, and chat history fresh in Nia with continuous synchronization. Start with Local Sync.

Connectors

Integrate external data sources through a unified framework with OAuth and API key authentication, scheduled syncing, and status monitoring. Start with Connectors.

Context Sharing

Save plans, discoveries, edited files, and conversation state so another agent can pick up where you left off. Use context.

Scoped MCP

Create a smaller, task-specific MCP surface for a particular source or workflow. Start with Scoped MCP Servers.

Explore Pre-indexed Knowledge

Ask questions across existing indexed sources without setting up everything yourself. Use Explore & Chat and Pre-indexed Sources.

​
Pick the Right Starting Point

If your goal is…Start herePrimary tools or pages
Bring a new source into NiaIndex & Subscribeindex, manage_resource, auto_subscribe_dependencies
Search or inspect what is already indexedSearch, Read & Exploresearch, nia_read, nia_grep, nia_explore
Search public package code without setupSearch Without Indexingnia_package_search_hybrid
Search public GitHub repos without setupSearch Without IndexingTracer
Deploy an agent to research a specific PDF or documentDocument AgentDocument Agent, POST /document/agent
Extract structured data from PDFsData ExtractionData Extraction, POST /extract
Compare options or investigate a questionResearch & Analysisnia_research, Oracle Research Agent
Review your code against docsResearch & Analysisnia_advisor
Connect cloud file storage and keep it syncedGoogle DriveGoogle Drive Integration
Integrate external data sourcesConnectorsConnectors
Keep local knowledge fresh over timeLocal SyncLocal Sync
Hand off work between agentsContext Sharingcontext
Reduce tool surface for a focused workflowScoped MCPScoped MCP Servers

​
Common Starting Flows

1

Bring in a Source

Index a repository, documentation site, paper, dataset, spreadsheet, or local folder with index, or use the dedicated Google Drive Integration flow for Drive content.
2

Search or Inspect It

Use search for semantic retrieval, nia_grep for pattern matching, nia_read to open content, and nia_explore to inspect structure.
3

Escalate to Research When Needed

Use nia_research, Tracer, or Oracle Research Agent when you need synthesis, comparisons, or deeper investigation.
4

Save What Matters

Use context to save the findings, then reuse them in another agent or follow-up session.

​
Tool Guide

​
index

Universal entry point for repositories, documentation, research papers, HuggingFace datasets, spreadsheets, and local folders.Auto-detects:
  • GitHub URLs as repositories
  • arXiv and PDF URLs as papers or PDFs
  • HuggingFace dataset URLs as datasets
  • CSV, TSV, XLSX, and XLS files as spreadsheets
  • Local paths as local folders
  • Other web URLs as documentation
Typical prompts:
"Index https://github.com/owner/repo"
"Index https://docs.example.com"
"Index https://arxiv.org/abs/2401.12345"
"Index https://huggingface.co/datasets/openai/gsm8k"

​
auto_subscribe_dependencies

Parse a manifest such as package.json, requirements.txt, pyproject.toml, Cargo.toml, or go.mod, then subscribe or index related documentation sources automatically.Best for:
  • dependency onboarding
  • spinning up a project knowledge base quickly
  • keeping documentation context aligned with a codebase

​
manage_resource

List, check status, rename, delete, or subscribe to indexed resources.Most common actions:
  • list
  • status
  • rename
  • delete
  • subscribe
Semantic search across indexed repositories, docs, papers, datasets, spreadsheets, Google Drive, and local folders.Best for:
  • answering natural-language questions
  • finding relevant sections before reading deeply
  • searching across multiple sources at once

​
nia_read

Read content from a repository, documentation source, package, Google Drive source, local folder, or HuggingFace dataset.

​
nia_grep

Regex search across repositories, documentation, packages, Google Drive sources, local folders, and datasets.

​
nia_explore

Browse trees and directories before reading individual files or rows.

​
get_github_file_tree

Inspect a GitHub repository structure without indexing it first.

​
nia_package_search_hybrid

Semantic search across public package source code without indexing first. Supports PyPI, npm, Crates.io, and Go modules.Typical prompts:
"Search the requests package for how error handling works"
"Find authentication examples in the fastapi package"

​
Document Agent

An autonomous AI agent you deploy into any indexed PDF or document. It plans its own research strategy and iteratively uses tools (search, read sections, read pages, navigate trees) to produce comprehensive answers with citations.Key features:
  • Section-aware citations with page numbers and section paths
  • Structured output via json_schema for programmatic extraction
  • Model selection: Opus 4.6 (1M context), Sonnet, etc.
  • Extended thinking for thorough analysis
  • Streaming support via SSE
Typical prompts:
"Query this SEC filing for all risk factors"
"Extract the methodology section from this paper"
"What are the key findings? Return as structured JSON"
→ Document Agent docs

​
Data Extraction

Extract structured records from PDFs using JSON schemas. Two modes: table extraction for tabular data, and engineering extraction for technical documents.Typical prompts:
"Extract all financial line items from pages 15-30"
"Parse the equations and specifications from this technical manual"
→ Data Extraction docs

​
nia_research

AI-powered research with three modes for different levels of depth.
ModeBest for
quickFast web search and quick source discovery
deepComparisons, evaluations, and multi-source analysis
oracleComplex multi-step investigations with autonomous research

​
nia_advisor

Analyze your code against indexed documentation for grounded recommendations.Typical prompts:
"Use nia_research deep to compare Prisma vs Drizzle"
"Use nia_research oracle to investigate authentication best practices for Next.js"
"Use nia_advisor to check my authentication code against the BetterAuth docs"
Related pages:

​
context

Save, retrieve, search, update, and delete cross-agent conversation context.Memory types:
  • scratchpad for short-lived working memory
  • episodic for session-level continuity
  • fact for persistent facts
  • procedural for reusable how-to knowledge
Typical prompts:
"Save this conversation about authentication setup"
"Search my saved contexts for database migrations"
This page is the clean capability view. If you are organizing docs around user inputs like code, docs, PDFs, datasets, Slack, or Google Drive, continue with Source Types.