π¦
Percept
Ambient voice intelligence for AI agents. Wearable mics to agent actions via MCP.
0 installs
Trust: 37 β Low
Other
Ask AI about Percept
Powered by Claude Β· Grounded in docs
I know everything about Percept. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Loading tools...
Reviews
Documentation
Percept β The Context Layer for AI Agents
Give your AI agent ears, eyes, and awareness.
Percept connects your real-world data to your AI agent through hardware capture, tool connectors, and a knowledge graph that enables your agent to take initiative.
What It Does
- π§ Hear β Omi pendant captures ambient audio. Chrome extension captures browser audio. Meetings, conversations, podcasts β all transcribed and searchable.
- β Interact β Apple Watch app with push-to-talk, raise-to-speak, and complications.
- π Connect β Gmail, GitHub, Linear, Calendar, Slack connectors sync your work context.
- π§ Understand β Knowledge graph connects people, projects, conversations, and events across all sources.
- β‘ Act β Initiative engine detects patterns and triggers actions without being asked.
Quick Start
pip install getpercept
percept serve # Start the audio pipeline
percept sync # Sync all connected tools
percept status # Check what your agent knows
Architecture
Hardware (Omi, Watch, Chrome) β Audio Pipeline β Transcription
Tool Connectors (Gmail, GitHub, Linear) β Data Sync
β
Knowledge Graph (entities, relationships, temporal)
β
Initiative Engine (patterns β actions)
β
Your AI Agent (OpenClaw, Claude, GPT, LangChain)
Package Structure
percept/
βββ percept/ # Python package
β βββ audio/ # Audio pipeline (receiver, transcriber, context)
β βββ core/ # Knowledge graph (graph DB, ingest, query, temporal)
β βββ connectors/ # SDK + Gmail, GitHub, Linear, Calendar, Slack
β βββ pipeline/ # Orchestration (sync β KG β signals β initiatives)
β βββ initiatives/ # Rules engine, pattern matching, actions
β βββ mesh/ # Team agent shared context
β βββ memory/ # Entity extraction, FTS5 search
β βββ mcp/ # MCP server for Claude Desktop
β βββ cli.py # Unified CLI entry point
βββ src/ # Legacy audio module (original package root)
βββ watch-app/ # Apple Watch app (Swift)
βββ extension/ # Chrome extension (browser audio capture)
βββ web/ # Landing page
βββ docs/ # Documentation
βββ tests/ # Test suite
Connectors
| Connector | Status | What It Captures |
|---|---|---|
| Omi Pendant | β Live | Ambient audio, meetings, conversations |
| Apple Watch | β TestFlight | Push-to-talk, raise-to-speak |
| Chrome Extension | β Built | Browser tab audio (meetings, YouTube, podcasts) |
| Gmail | β Live | Emails, threads, contacts |
| GitHub | β Live | PRs, issues, commits, reviews |
| Linear | β Live | Tickets, projects, team activity |
| Calendar | πΆ Ready | Events, attendees (needs Google API enabled) |
| Slack | π Planned | Messages, channels, threads |
CLI Commands
percept sync # Run all connectors β KG β initiatives
percept sync --connector gmail # Single connector
percept status # KG stats + connector health + initiatives
percept query "..." # Search the knowledge graph
percept serve # Start the audio pipeline + API
percept connectors # List installed connectors
percept initiatives # List triggered initiatives
Works With Any AI Framework
Percept is framework-agnostic. It provides context via:
- CLI β
percept sync,percept query - MCP Server β 8 tools for Claude Desktop integration
- REST API β HTTP endpoints for any framework
- Python SDK β
from percept.core.graph.database import GraphDB
License
MIT
