alvin-bot
Alvin Bot â Your personal AI agent on Telegram, WhatsApp, Discord, Signal, and Web.
Why this rank:Recent releaseHealthy release cadence
Description
Alvin Bot â Your personal AI agent on Telegram, WhatsApp, Discord, Signal, and Web.
Release History
| Version | Changes | Urgency | Date |
|---|---|---|---|
| v4.25.1 | ## Fix: silent PM2 â launchd migration leak If you migrated alvin-bot from PM2 to launchd at any point in the past, PM2 may have been silently resurrecting alvin-bot on every login from a stale `~/.pm2/dump.pm2` â fighting your launchd-managed bot for the same Telegram `BOT_TOKEN` and surfacing eventually as `409 Conflict: terminated by other getUpdates request` errors. This release closes that gap. ### Root cause `alvin-bot launchd install` already called `pm2 delete alvin-bot` to remove the | High | 5/13/2026 |
| v4.19.2 | ## đ Fixes ### Workspace switch produced `(no response)` cascade After v4.19.1 shipped, a `/workspace <name>` or `/dir <path>` switch could still produce broken responses â but this time NOT empty streams. Claude replied with literal text like `(no response)\n\nUser: Hallo`, and each follow-up turn got worse. **Root cause.** v4.19.1's cwd-change reset set `session.lastSdkHistoryIndex = -1`. That value is consumed by `buildBridgeMessage()` in `handlers/message.ts`, which is designed for the O | High | 4/24/2026 |
| 4.18.1 | Imported from npm (4.18.1) | Low | 4/21/2026 |
| v4.18.1 | Adds automated pre-publish PII scan. See CHANGELOG for details. | High | 4/20/2026 |
| v4.18.1 | Adds automated pre-publish PII scan. See CHANGELOG for details. | High | 4/20/2026 |
| v4.18.1 | Adds automated pre-publish PII scan. See CHANGELOG for details. | High | 4/20/2026 |
| v4.18.0 | Completes the audit work from 4.17.0. **Performance (hot path):** - User profiles cached in memory + debounced writes (no more 2 sync fs ops per message) - Embeddings index cached with mtime invalidation (previously re-read on every query) - Skills no longer force-reload every 5min (fs.watch already handles hot-reload) **Hardening:** - Sub-agents map capped at 1000 with oldest-first eviction of delivered entries - Async-agent pending map capped at 500 - Browser gateway + MCP subprocess stderr | High | 4/20/2026 |
| v4.17.0 | ## Hardening: long-running stability audit + leak fixes **Fixes:** - WhatsApp event-listener leak on reconnect (memory + duplicate messages) - CDP file-descriptor leak on Chromium bootstrap - Heartbeat + auto-update timers now properly cleaned up on shutdown ## New: /cleanup feature Automatic daily cleanup of transient files (logs, screenshots, old subagent outputs, temp media). Memory, assets, cron jobs and workspaces stay untouched. - `/cleanup` â show policy - `/cleanup run` â trigger man | High | 4/20/2026 |
| v4.16.1 | After a successful `/update`, the bot now sends a short summary of what changed in the new version. Pulled from CHANGELOG.md, max 5 bullets, Telegram-friendly. See CHANGELOG.md for details. | High | 4/20/2026 |
| v4.16.0 | ## Feature: bot-owned CDP Chromium â no more hub dependency **New:** - `src/services/cdp-bootstrap.ts` â spawns Playwright's bundled Chromium (no conflict with user Chrome) - `alvin-bot browser` CLI subcommand â start/stop/goto/shot/eval/tabs/doctor, with self-healing - `browser-manager` rewired â CDP strategy works on every install, no external dependency **First-run setup:** `npx playwright install chromium` **Note:** All versions below 4.16.0 have been deprecated as part of a privacy/PII c | High | 4/20/2026 |
Dependencies & License Audit
Loading dependencies...
Similar Packages
fn-aiAI agent toolkit with pluggable provider, tool, and session store layers. ESM, TypeScript-first.0.6.0
@avasis-ai/synthcodeSynthCode - Synthesize any LLM into a production-grade AI agent. Battle-tested agentic patterns, model-agnostic, TypeScript-first.main@2026-05-09
@clanklabs/clankLocal-first AI agent harness. Open-source alternative to OpenClaw, optimized for local models.v1.12.2
More in AI Agents
hermes-agentThe agent that grows with you
awesome-copilotCommunity-contributed instructions, agents, skills, and configurations to help you make the most of GitHub Copilot.
e2bE2B SDK that give agents cloud environments
letta-codeThe memory-first coding agent