apteva
AI agent platform with multi-thread orchestration, MCP integrations, and dashboard
Description
AI agent platform with multi-thread orchestration, MCP integrations, and dashboard
Release History
| Version | Changes | Urgency | Date |
|---|---|---|---|
| 0.9.5 | Imported from npm (0.9.5) | Low | 4/21/2026 |
| v0.9.5 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.9.4...v0.9.5 | High | 4/19/2026 |
| v0.9.4 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.9.3...v0.9.4 | High | 4/19/2026 |
| v0.9.3 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.9.2...v0.9.3 | High | 4/18/2026 |
| v0.9.2 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.9.1...v0.9.2 | High | 4/18/2026 |
| v0.9.1 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.9.0...v0.9.1 | High | 4/18/2026 |
| v0.9.0 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.8.4...v0.9.0 | High | 4/18/2026 |
| v0.8.4 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.8.3...v0.8.4 | Medium | 4/11/2026 |
| v0.8.3 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.8.2...v0.8.3 | Medium | 4/11/2026 |
| v0.8.2 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.8.1...v0.8.2 | Medium | 4/11/2026 |
| v0.8.1 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.8.0...v0.8.1 | Medium | 4/11/2026 |
| v0.8.0 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.6...v0.8.0 | Medium | 4/11/2026 |
| v0.7.6 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.5...v0.7.6 | Medium | 4/11/2026 |
| v0.7.5 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.4...v0.7.5 | Medium | 4/10/2026 |
| v0.7.4 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.3...v0.7.4 | Medium | 4/10/2026 |
| v0.7.3 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.2...v0.7.3 | Medium | 4/10/2026 |
| v0.7.2 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.7.0...v0.7.2 | Medium | 4/10/2026 |
| v0.7.1 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.5.0...v0.7.1 | Medium | 4/10/2026 |
| v0.7.0 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.5.0...v0.7.0 | Medium | 4/10/2026 |
| v0.5.0 | ## Apteva v0.5.0 Complete rewrite — Go-based continuous thinking agents with multi-thread coordination. ### Highlights - **Multi-thread agents** — main coordinator spawns worker threads with dedicated tools and directives - **Telemetry pipeline** — full event streaming (tool calls, results, thread messages) to dashboard and DB - **Webhook subscriptions** — auto-register webhooks with external services (Stripe, OmniKit, etc.) - **265 integrations** — connect to any service via MCP with encrypte | Medium | 3/31/2026 |
| v0.4.57 | ## What's New - **Live config push** — `update_agent` now pushes config to running agents automatically. No restart needed for feature changes, system prompt updates, skill/MCP assignments, etc. - **Clearer MCP responses** — `create_agent` and `update_agent` return human-readable feature names (e.g. "multi-agent (call_agent, delegate_task, list_available_agents)" instead of "agents") - Response includes `config_status: "applied live (no restart needed)"` when the agent is running **Full Change | Low | 3/10/2026 |
| v0.4.56 | ## What's New - **Multi-agent MCP support** — `create_agent` and `update_agent` platform MCP tools now expose the `agents` feature flag, enabling the meta agent to set up multi-agent communication - **Test mode header** — Test runner sends `X-Test-Mode: true` to agents, activating tool mocking and memory extraction bypass - **Test history drawer** — Run history opens as a right-side drawer panel instead of inline, matching the agent panel UX ## Other Changes - UI mode context, provider update | Low | 3/10/2026 |
| v0.4.53 | ## Changes - Bump agent binary to v1.44.9 (was pinned to v1.44.6 via lockfile) Users running `npx apteva@latest` will now get agent v1.44.9 out of the box. **Full Changelog**: https://github.com/apteva/apteva/compare/v0.4.52...v0.4.53 | Low | 3/2/2026 |
| v0.4.52 | ## What's New ### Ollama Provider Support - Full local Ollama integration — run agents with local models - Dynamic model detection from running Ollama instance - No API key required for local providers ### UI Improvements - Renamed "Telemetry" → "Analytics" across all pages - Token usage chart now shows stacked bars with cache read/write and reasoning token breakdown - Charts update in real-time via SSE (no refresh needed) ### Fixes - Fixed auth for Ollama status/model endpoints (was showing | Low | 3/2/2026 |
| v0.4.51 | ## Changes - Rename Telemetry page to Analytics across all UI - Token usage chart now shows stacked breakdown (cache read/write, reasoning) with orange/purple tints - Charts now update in real-time via SSE (merged fetched + live data) - Fix migration ordering: 024_create_test_runs now runs before 026/028 that ALTER it (fixes fresh installs) - Bump agent binary dependency to ^1.44.6 for latest agent on fresh installs **Full Changelog**: https://github.com/apteva/apteva/compare/v0.4.41...v0.4.51 | Low | 3/2/2026 |
| v0.4.48 | - Upgrade @types/react and @types/react-dom to v19 (fixes npm update peer dependency conflicts) - Cache token telemetry support (cache_creation_tokens, cache_read_tokens) - Reasoning token telemetry support - Agent version fix: write VERSION file to agent data dir so npm-installed binaries report correct version - Professional theme font switching (Inter) - Telemetry stats cards single-row layout - Settings color scheme card border fix | Low | 2/28/2026 |
| v0.4.41 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.4.32...v0.4.41 | Low | 2/22/2026 |
| v0.4.32 | **Full Changelog**: https://github.com/apteva/apteva/compare/v0.4.31...v0.4.32 | Low | 2/16/2026 |
| v0.4.30 | ## What's New ### Browser Provider Support (Operator Mode) - **Browserbase, Steel, BrowserEngine, Chrome** added as configurable browser providers in Settings UI - Two-field UI for Browserbase (API Key + Project ID) — stored as encrypted JSON - Browser provider picker in agent creation and settings panel - Operator config resolves browser credentials per project ### CDP Connection Reliability - **Auto-reconnect** on WebSocket disconnect — retries once before failing - **Browserbase**: `keepAli | Low | 2/14/2026 |
| v0.4.20 | ## What's New ### MCP Subscription Tools - 8 new MCP tools for agents to programmatically manage trigger subscriptions (list providers, create/enable/disable/delete subscriptions, list connected accounts) - Supports all trigger providers (Composio, AgentDojo) ### Project-Scoped Data Loading - Agents, dashboard stats, tasks, and telemetry now filter server-side by project - Eliminates unnecessary data transfer when project mode is enabled ### Faster Page Load - Merged onboarding status into au | Low | 2/12/2026 |
| v0.4.19 | ## What's New ### AgentDojo Trigger Subscriptions - Create subscriptions directly from the AgentDojo triggers tab - Dynamic config fields rendered from trigger type's `config_schema` (e.g. owner/repo for GitHub webhooks) - Auto-create local subscription record for webhook routing ### Fixes - Fixed AgentDojo API URL doubling (`/mcp/mcp` → `/mcp`) - Fixed provider key lookup to use project-scoped resolution - Pass extra config fields (owner, repo, etc.) through to MCP webhook register functions | Low | 2/11/2026 |
| v0.4.15 | ## What's New ### Built-in MCP Platform Server 🤖 Agents can now control the Apteva platform itself through a built-in MCP server - manage other agents, MCP servers, skills, and integrations programmatically via a meta agent. - **Platform Control Tools** - Start/stop agents, manage MCP servers and skills, all from within an agent conversation - **Enriched Tool Descriptions** - MCP tool descriptions automatically enriched with provider/model/feature documentation - **Skills & MCP Management Too | Low | 2/8/2026 |
| v0.3.0 | ## What's New ### Skills Support 🎯 Agents can now use **Skills** - reusable instruction sets that can be shared across multiple agents. #### Features - **Skills Page** - New dedicated page for managing skills - **Marketplace Integration** - Browse and install skills from GitHub (anthropics/skills repository) - **Local Skills** - Create custom skills with markdown content - **Agent Integration** - Attach skills to agents in Settings tab - **Auto-sync** - Skill updates automatically push to ru | Low | 2/3/2026 |
Dependencies & License Audit
Loading dependencies...
Similar Packages
automagik-genieSelf-evolving AI agent orchestration framework with Model Context Protocol support2.5.27
council-mcpThree-tier AI agent MCP orchestration system: Chancellor (Opus), Executor (Sonnet), Aide (Haiku)0.4.0
@avasis-ai/synthSynthesize any LLM into a production-grade AI agent. Battle-tested agentic patterns, model-agnostic, TypeScript-first.0.6.0