█████╗ ██████╗ ██╗███████╗ █████╗ ██╔══██╗██╔══██╗██║██╔════╝██╔══██╗ ███████║██████╔╝██║███████╗███████║ ██╔══██║██╔══██╗██║╚════██║██╔══██║ ██║ ██║██║ ██║██║███████║██║ ██║ ╚═╝ ╚═╝╚═╝ ╚═╝╚═╝╚══════╝╚═╝ ╚═╝
Arisa: an open-source AI assistant for Telegram
Run it on your own machine. Powered by Pi Agent.
1.1about
Arisa keeps exactly two things and nothing more: Telegram as the human interface and Pi Agent as the reasoning engine.
No pre-loaded opinions about what the agent should do or which tools it should have. It starts minimal and grows from real use.
- Telegram is the human interface
- Pi Agent is the reasoning engine
- everything is an artifact
- capabilities live in isolated CLI tools
- tools can be chained through pipes
1.2load
2install
Install globally, then run arisa. First launch guides you through Telegram setup, Pi provider/model selection, authentication, and validation.
ChatGPT subscription is enough: Arisa can authenticate through Pi's Codex OAuth flow, so you do not need a standard OpenAI API key when using supported providers.
3arch
Arisa separates pre-reasoning normalization from reasoned action. Voice can become text before reasoning; TTS, downloads, and other chains happen after the agent decides.
4zero tools
A fresh install ships with zero tools. The core is Telegram transport, the Pi Agent reasoning loop, the artifact store, and the tool registry.
- a request arrives that current tools cannot satisfy
- Arisa checks the official catalog for a matching tool
- low-footprint tools install immediately; heavier tools ask for confirmation
- if no catalog tool fits, Arisa builds the missing capability
- installed tools live in ~/.arisa/tools/ and are reused later
understand
Search, read, and retain useful context.
media
Download, inspect, transcribe, and speak.
create and publish
Turn text and ideas into files or live sites.
communicate
Work across email, WhatsApp, and X with explicit safety rules.
operate
Schedule work, connect machines, and run realtime infrastructure.
extend and maintain
Connect external tools, review operations, and keep installations reproducible.
5cookbook
Arisa is not a feature list. It adapts to the work: it connects tools, keeps context, and lets each chat decide how much control to retain.
WebRTC without a game serversignaling-server coordinates peers and turn-server provides TURN relays when a direct connection is not enough. Useful for rooms, calls, data channels, and multiplayer games.
idea → design → HTTPSAsk for a landing page or internal tool: site-builder shapes the page and roster-sites publishes it with a domain and HTTPS.
secure profile → specialized actionmcp-client connects chat-scoped MCP profiles with guarded OAuth and endpoint validation. magnific-mcp turns that transport into direct image generation, collection, download, and upscale workflows.
master → paired slave → bounded operationmaster-slave pairs a main Arisa with headless Linux Slaves. Each Slave reconnects outward and exposes only the roots and inspect, read, or command capabilities allowed by its stored policy.
observe → explain → proposeprocess-retrospective reviews recent operations on a rotating focus, surfaces a bounded set of improvements, and never applies changes without approval.
group, lookup, or replywhatsapp-web keeps one session per chat. It can reply in a group, search the web to compare prices, or send received audio through openai-transcribe before Arisa replies.
research → draft → Gmaillightpanda-browser finds a real reference. pr-campaign tracks the contact, creates the personalized outreach, and calls gmail-workspace to save it in Gmail Drafts. In review mode, it waits for approval; in automatic mode, it sends only within the campaign rules. stop-slop-writer checks the tone.
discover → approve → verifyx-session-reader reads public session content, x-campaign-runner selects one evidence-backed prospect, and x-dm follows or sends only after explicit approval and account checks.
context → scheduled taskcontext-vault stores chat-scoped facts and preferences. schedule-agent-task wakes the agent later with the right instruction.
web page → text → voiceurl-to-markdown extracts readable text from a page, then openai-tts returns an OGG voice message.
research → Markdown → PDFweb-browser gathers sources, markdown-pdf renders the report, and file-document can return an existing local file through Telegram.
authorized auditweb-pentest-auditor checks headers, CORS, public routes, bundles, and token storage without attacking or modifying the site. It is technical review a normal chat cannot run on its own.
download → inspect → transcribevideo-downloader fetches the source, video-analyze samples visual evidence, and audio-extractor prepares sound for openai-transcribe or local whispermix-transcribe.
one tool prepares the nextTools do not live in isolation: URL → video → audio → transcript; research → reference → draft; message → WhatsApp reply. Arisa assembles the pipe it needs and reuses it.
6philosophy
Arisa starts small and changes through the work you give it.
Telegram carries the conversation. Pi Agent handles the reasoning. Arisa adds each new capability as an isolated tool when someone needs it.
In Stories, Arisa documents how those choices shape real work.
Read Stories →- 01 · user-shaped
- The person using Arisa chooses what it may access, remember, automate, and publish.
- 02 · tool-shaped
- Arisa puts each new capability in an isolated tool. The core stays focused on transport, reasoning, artifacts, and orchestration.
- 03 · inspectable
- Arisa records artifacts, state, or receipts for its work and asks for approval before sensitive changes.
- 04 · self-improving
- Arisa can read and patch its own code. The person using it decides which changes remain.