Mayo CLI Documentation

Mayo CLI is an autonomous AI companion for your terminal, designed to transform how you interact with data. Whether you're a data scientist, researcher, or developer, Mayo helps you query, analyze, and enrich your datasets using natural language — all while keeping your data private and secure.

Installation

curl -fsSL https://mayo.teleskop.id/install.sh | bash

Termux (Android)

bash install_termux.sh

Prerequisites

• SQLite3 — Required for local data storage and in-memory processing.
• Go 1.21+ — Only if building from source.
• AI Provider API Key — At least one: OpenAI, Google Gemini, Anthropic, or Groq.

First Run

After installation, start Mayo and run the onboarding wizard:

mayo
> /wizard

The wizard guides you through:

1. Selecting and configuring your AI Provider (API key, model).
2. Connecting your first data source (database or file).
3. Setting up privacy preferences (PII masking, interactive mode).

All Commands

Mayo operates in an interactive shell. Use slash commands (/command) for administrative tasks and natural language for data queries.

AI & Configuration

Manage AI behavior, privacy settings, and persistent context.

Connectivity & Data Sources

Connect Mayo to your databases and files.

Data Processing (Dataframes)

Mayo uses "Dataframe Mode" to stage results in memory for complex analysis.

Knowledge Base (RAG)

Build a semantic memory from external documents.

Sessions & Research

Organize your research threads with persistent sessions.

API & System

Terminal Commands (External)

Commands run directly from your shell, not inside Mayo's interactive mode.

Natural Language Queries

Mayo's core feature is the ability to ask questions about your data in plain English. Here's how it works:

1. Connect — Link a database or file with /connect.
2. Ask — Type your question naturally. Mayo generates SQL based on your schema metadata.
3. Review — With Interactive Mode enabled, review the generated SQL before execution.
4. Results — Data is fetched, displayed as a table, and optionally analyzed by the AI analyst.

mayo▸ What are the top 10 customers by total spend this quarter?

 Generating SQL...
SELECT customer_name, SUM(amount) as total_spend
FROM orders
WHERE order_date >= '2026-01-01'
GROUP BY customer_name
ORDER BY total_spend DESC
LIMIT 10;

✓ Run this query? [Y/n]: y
✓ 10 rows returned in 0.34s

Data Enrichment

The AI Data Enhancer processes your datasets row-by-row using an AI model. Common use cases:

• Sentiment Analysis — Classify customer feedback as positive/negative/neutral.
• Data Classification — Categorize products, transactions, or text entries.
• Data Cleaning — Standardize names, addresses, or fix formatting issues.
• Content Generation — Generate descriptions, titles, or summaries from raw data.

/enhance start --prompt "Classify the sentiment of the 'feedback' column"

The enhancer runs incrementally and can be stopped and resumed. Progress is saved in your local SQLite database.

Data Reconciliation

The /reconcile command compares data across two connected sources, identifying discrepancies with AI assistance.

/reconcile
> Select source A: postgres (analytics_db)
> Select source B: file (bank_statement.csv)
> AI analyzing mapping patterns...

Mayo presents a summary of findings, then lets you choose between automatic reconciliation or manual row selection (git-commit style).

Knowledge Base (RAG)

Build a semantic memory from your documents for contextual queries.

/knowledge rfp_proposal.pdf
✓ Indexed 247 chunks from rfp_proposal.pdf

mayo▸ What are the compliance requirements mentioned in the RFP?

You can also cross-reference knowledge documents against each other for gap analysis:

/knowledge against compliance_checklist.md

API Server

Expose Mayo as a REST API for integration with other applications.

mayo serve --port 8080

Endpoints available:

Trust & Security

Mayo is designed with a "Security First, Privacy Always" philosophy. Here are the key security pillars:

PII Masking

Mayo's built-in PII Detection Engine (internal/privacy) automatically scans for sensitive information before any cloud communication:

• Emails → Replaced with «EMAIL_1»
• Phone Numbers → Replaced with «PHONE_1»
• Credit Cards → Replaced with «CC_1»

Tokens are mapped back to original values only on your local machine. Toggle with /privacy.

Prompt Auditing

Use /audit to see the exact JSON payload sent to the AI at any time. Audit logs are stored at:

~/.mayo-cli/sessions/[session_id]/audit.log

Secure Credentials

API keys are stored in your OS Keychain (macOS Keychain / Windows Credential Manager) — never in plain text.

Network Verification

Mayo only communicates with your database and your chosen AI provider. Verify using network tools like Little Snitch or Wireshark.

Architecture

• Core — Written in Go for high performance and easy distribution as a single binary.
• Engine — Custom Orchestrator for LLM communication and safe SQL execution.
• UI — Powered by lipgloss and charmbracelet for a premium terminal experience.
• Storage — Local metadata and session state via SQLite and JSON files.
• Privacy — PII detection, secure keychain integration, and transparent auditing.

Changelog

View the latest feature updates directly in Mayo:

/changelog

Or generate AI-powered release notes from git commits:

mayo release-notes