titan-agent
TITAN â Autonomous AI agent framework with self-improvement, multi-agent orchestration, 36 LLM providers, 16 channel adapters, GPU VRAM management, mesh networking, LiveKit voice, TITAN-Soma homeostat
Why this rank:Recent releaseStrong adoptionHealthy release cadence
Description
TITAN â Autonomous AI agent framework with self-improvement, multi-agent orchestration, 36 LLM providers, 16 channel adapters, GPU VRAM management, mesh networking, LiveKit voice, TITAN-Soma homeostatic drives, and a React Mission Control dashboard. Open-
Release History
| Version | Changes | Urgency | Date |
|---|---|---|---|
| v7.2.1 | # TITAN v7.2.1 â Reliability Mode composes all four Fable-5 disciplines Building on v7.2.0 "Conscience", one switch (`agent.reliabilityMode: true`) now gives a local model the **complete disciplined-process nervous system** of a frontier agent: 1. **Plan-first** â on multi-step tasks, write a numbered plan and verify each step 2. **Verify** â never claim a side-effect without the tool that did it (the wall) 3. **Self-critique** â review your own draft adversarially, append honest caveats 4. ** | Medium | 7/7/2026 |
| v7.0.0 | # TITAN v7.0.0 "Independence" đ **Local-first is independence: your AI, your hardware, your data.** ## Headliners - đĒ **Muscle Memory** â the first trustworthy automatic self-improvement in any agent framework. TITAN notices your repeated workflows, teaches itself a parameterized skill, and **proves it works by replaying it against your real usage through a deterministic eval harness** before you ever see it. One click to adopt â instant slash command. Nothing auto-adopts, side-effectful wo | High | 7/3/2026 |
| v5.4.1 | Patch release. Tightens the provider, UI, mesh, and config layers around the v5.4.0 framework so every transient failure has a predictable, testable recovery path. **1057 deterministic tests pass in 19s** across 41 files. Typecheck clean. Created by Tony Elliott aka djtony707. ### Added **Per-model output clamping (Kimi)** - New `src/providers/modelCapabilities.ts` â central source of truth for every provider's context-window + max-output ceiling, plus a family- pattern fallback for unkn | High | 4/27/2026 |
| v5.0.1 | # đ v5.0.1 â Spacewalk The biggest TITAN release since v1.0. Mission Control becomes a browser-first widget canvas, 110 production-ready widget templates ship out of the box, multi-agent orchestration is hardened, observability is end-to-end, and the agent gains live situational awareness of the canvas it's running in. ## Install ```bash npm install -g titan-agent@next titan onboard titan gateway ``` After ~1 week of real-world feedback, `5.0.1` promotes from `@next` â `@latest`. ## Headli | High | 4/26/2026 |
| 4.13.0 | Imported from npm (4.13.0) | Low | 4/21/2026 |
| v1.0.0 | ## TITAN v1.0.0 **TITAN goes semver.** This is the first stable release under proper semantic versioning, replacing the `2026.10.XX` date-based scheme. All prior versions are deprecated. ### Install ```bash npm install titan-agent@1.0.0 ``` ### Highlights - **Paperclip Integration** â Full agent governance: types, API client, routes, and 14 Command Post UI components - **Provider Error Recovery** â Circuit breaker pattern, exponential backoff retry, automatic fallback chain - **Multi-Agent A | Medium | 4/4/2026 |
| v1.0.0 | ## TITAN v1.0.0 **TITAN goes semver.** This is the first stable release under proper semantic versioning, replacing the `2026.10.XX` date-based scheme. All prior versions are deprecated. ### Install ```bash npm install titan-agent@1.0.0 ``` ### Highlights - **Paperclip Integration** â Full agent governance: types, API client, routes, and 14 Command Post UI components - **Provider Error Recovery** â Circuit breaker pattern, exponential backoff retry, automatic fallback chain - **Multi-Agent A | Medium | 4/4/2026 |
| v1.0.0 | ## TITAN v1.0.0 **TITAN goes semver.** This is the first stable release under proper semantic versioning, replacing the `2026.10.XX` date-based scheme. All prior versions are deprecated. ### Install ```bash npm install titan-agent@1.0.0 ``` ### Highlights - **Paperclip Integration** â Full agent governance: types, API client, routes, and 14 Command Post UI components - **Provider Error Recovery** â Circuit breaker pattern, exponential backoff retry, automatic fallback chain - **Multi-Agent A | Medium | 4/4/2026 |
| v1.0.0 | ## TITAN v1.0.0 **TITAN goes semver.** This is the first stable release under proper semantic versioning, replacing the `2026.10.XX` date-based scheme. All prior versions are deprecated. ### Install ```bash npm install titan-agent@1.0.0 ``` ### Highlights - **Paperclip Integration** â Full agent governance: types, API client, routes, and 14 Command Post UI components - **Provider Error Recovery** â Circuit breaker pattern, exponential backoff retry, automatic fallback chain - **Multi-Agent A | Medium | 4/4/2026 |
| v2026.10.49 | ## Hindsight MCP Bridge â Cross-Session Episodic Memory ### Added - **Hindsight MCP Bridge** (`src/memory/hindsightBridge.ts`) â Connects TITAN's learning system to Vectorize.io's Hindsight episodic memory - `retainStrategy()` â Stores successful tool sequences as "experience" memories after each task - `getHindsightHints()` â Queries cross-session memory when local strategy hints don't match - Fully fire-and-forget â never blocks or crashes if Hindsight is unavailable - 14 new tests co | Medium | 3/25/2026 |
| v2026.10.48 | ## Smart Auto-Learning ### Added - **SmartCompress Plugin** â Task-type-aware context compression via ContextEngine plugin. Classifies conversations by type (coding, research, analysis, general) and applies optimal compression strategies. Configurable aggressiveness. - **Continuous Learning Feedback Loop** â Strategy outcome tracking with `recordStrategyOutcome()`. Failed strategies auto-excluded from hints. 30-day decay for unvalidated strategies. ### Fixed - **Ordered Tool Sequence Capture** | Medium | 3/25/2026 |
| v2026.10.47 | ## What's New ### Multi-Chip GPU Support - **NVIDIA** (nvidia-smi), **AMD ROCm** (rocm-smi), and **Apple Silicon** (Metal/MPS) GPU detection and monitoring - Vendor-aware VRAM orchestrator dispatches to correct probe per platform - Apple Silicon unified memory support (GPU shares system RAM) - `system_info` and `vram_status` tools report vendor-specific GPU details - New config option: `vram.gpuVendor` (`auto` | `nvidia` | `amd` | `apple` | `none`) ### Hindsight MCP Memory Preset - Builtin MCP | Medium | 3/25/2026 |
| v2026.10.43 | ## What's New ### VRAM Orchestrator (v2026.10.43) Automatic GPU VRAM management for NVIDIA GPUs (tested on RTX 5090, 32GB). - **nvidia-smi polling** â periodic GPU state monitoring (VRAM usage, temp, utilization) - **Model auto-swap** â when GPU services need VRAM, the orchestrator downgrades the LLM to a smaller model, then upgrades back when done - **Time-bounded leases** â services reserve VRAM with auto-expiry, preventing memory hogging - **Async mutex** â all VRAM operations serialized, n | Low | 3/17/2026 |
| v2026.10.39 | ## đ Security Release Resolved **all 23 Dependabot vulnerability alerts** â TITAN now has **0 known vulnerabilities**. ### Changes - Upgraded `matrix-js-sdk` v34 â v41 - Added npm overrides for transitive dependencies: `esbuild` ^0.25.0, `yauzl` ^3.2.1, `langsmith` ^0.5.0 - All 3,879 tests passing across 124 test files ### Previous Release (v2026.10.38) - `titan doctor --json` â Machine-readable JSON output (Issue #2) - Better provider error messages with actionable hints (Issue #3) - npm do | Low | 3/16/2026 |
| v2026.10.37 | ## What's New ### đī¸ Streaming Voice Pipeline - LLM â sentence-chunked TTS with ~2s latency savings - Phase 2 of voice latency optimization ### đ¤ TITAN Agent Framework - **152 tools** across 93 skills - **34 providers** (4 native + 30 OpenAI-compatible) - **3,839+ tests** passing - LiveKit WebRTC voice integration - MCP Server mode (expose tools to other agents) - RAG with FTS5 + embeddings - SSE + WebSocket streaming - Mission Control v2 React dashboard - Mesh networking with peer approval | Low | 3/16/2026 |
Dependencies & License Audit
Loading dependencies...
Similar Packages
@stevederico/dotbotAI agent CLI and library for Node.js â streaming, multi-provider, tool execution, autonomous tasksmaster@2026-07-24
More in MCP Servers
supersetCode Editor for the AI Agents Era - Run an army of Claude Code, Codex, etc. on your machine
kreuzbergA polyglot document intelligence framework with a Rust core. Extract text, metadata, images, and structured information from PDFs, Office documents, images, and 91+ formats. Available for Rust, Python
ai-engineering-from-scratchLearn it. Build it. Ship it for others.
CodeGraphContextAn MCP server plus a CLI tool that indexes local code into a graph database to provide context to AI assistants.