open-agents-ai 0.185.70 → 0.185.71

package/README.md

@@ -1,3 +1,4 @@
+<a name="top"></a>
 <h1 align="center">Open Agents</h1>
 
 <p align="center">
@@ -24,6 +25,8 @@ An autonomous multi-turn tool-calling agent that reads your code, makes changes,
 
 ## Table of Contents
 
+<div align="right"><a href="#top">back to top</a></div>
+
 - [The Organism, Not the Cortex](#the-organism-not-the-cortex)
 - [How It Works](#how-it-works)
 - [Features](#features)
@@ -63,6 +66,8 @@ An autonomous multi-turn tool-calling agent that reads your code, makes changes,
 
 ## The Organism, Not the Cortex
 
+<div align="right"><a href="#top">back to top</a></div>
+
 An LLM is a high-bandwidth associative generative core — closer to a cortex-like prior than to a complete agent. Its weights contain broad latent structure, but they do not by themselves give you situated continuity, durable task state, calibrated action policies, or grounded memory management. Open Agents treats the model as one organ inside a larger organism. The framework provides the rest: sensors, effectors, memory stores, routing, gating, evaluation, and persistence.
 
 **What the framework provides:**
@@ -83,6 +88,8 @@ Don't chase larger models. Build the organism around whatever model you have.
 
 ## How It Works
 
+<div align="right"><a href="#top">back to top</a></div>
+
 ```
 You: oa "fix the null check in auth.ts"
 
@@ -97,6 +104,8 @@ The agent uses tools autonomously in a loop — reading errors, fixing code, and
 
 ## Features
 
+<div align="right"><a href="#top">back to top</a></div>
+
 - **61 autonomous tools** — file I/O, shell, grep, web search/fetch/crawl, memory (read/write/search), sub-agents, background tasks, image/OCR/PDF, git, diagnostics, vision, desktop automation, browser automation, temporal agency (scheduler/reminders/agenda), structured files, code sandbox, transcription, skills, opencode delegation, cron agents, nexus P2P networking + x402 micropayments, **COHERE cognitive stack** (persistent REPL, recursive LLM calls, memory metabolism, identity kernel, reflection, exploration)
 - **Moondream vision** — see and interact with the desktop via Moondream VLM (caption, query, detect, point-and-click)
 - **Desktop automation** — vision-guided clicking: describe a UI element in natural language, the agent finds and clicks it
@@ -199,6 +208,8 @@ D8AgCTrxpDKD5meJ2bpAfVwcST3NF3EPuy9xczYycnXn
 
 ## Enterprise & Headless Mode
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Run Open Agents as a headless service for CI/CD pipelines, automation, and enterprise deployments.
 
 ### Non-Interactive Mode
@@ -622,6 +633,8 @@ Free for non-commercial use under CC-BY-NC-4.0. For enterprise/commercial licens
 
 ## Architecture
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The core is `AgenticRunner` — a multi-turn tool-calling loop with structured context assembly:
 
 ```
@@ -642,6 +655,8 @@ User task → assembleContext(c_instr, c_state, c_know) → LLM → tool_calls
 
 ## Context Engineering
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The agent implements structured context assembly based on current research in context engineering, modular prompt optimization, and instruction hierarchy:
 
 ```
@@ -666,6 +681,8 @@ Research provenance: grounded in "A Survey of Context Engineering for LLMs" (con
 
 ## Model-Tier Awareness
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Open Agents classifies models into three tiers and adapts its behavior accordingly:
 
 | Tier | Parameters | Base Tools | System Prompt | Compaction |
@@ -701,6 +718,8 @@ All context-dependent values scale automatically with the actual context window
 
 ## Auto-Expanding Context Window
 
+<div align="right"><a href="#top">back to top</a></div>
+
 On startup and `/model` switch, Open Agents detects your RAM/VRAM and creates an optimized model variant:
 
 | Available Memory | Context Window |
@@ -714,6 +733,8 @@ On startup and `/model` switch, Open Agents detects your RAM/VRAM and creates an
 
 ## Tools (61)
 
+<div align="right"><a href="#top">back to top</a></div>
+
 | Tool | Description |
 |------|-------------|
 | **File Operations** | |
@@ -820,6 +841,8 @@ The agent has 4 web tools. Pick the right one:
 
 ## Ralph Loop — Iteration-First Design
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The Ralph Loop is the core execution philosophy: **iteration beats perfection**. Instead of trying to get everything right on the first attempt, the agent executes in a retry loop where errors become learning data rather than session-ending failures.
 
 ```
@@ -844,6 +867,8 @@ The loop tracks iteration history, generates completion reports saved to `.aiwg/
 
 ## Task Control
 
+<div align="right"><a href="#top">back to top</a></div>
+
 ### Pause, Stop, Resume, Destroy
 
 | Command | Behavior |
@@ -883,6 +908,8 @@ Type `y` to restore — the previous session context will be prepended to your n
 
 ## COHERE Cognitive Framework
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Open Agents implements the **COHERE layered cognitive stack** — a provenance-grounded architecture for persistent, reflective agentic systems. Each layer adds a distinct cognitive capability, grounded in specific research papers:
 
 ```
@@ -961,6 +988,8 @@ The identity kernel maintains a persistent self-model across sessions, the refle
 
 ## Context Compaction — Research-Backed Memory Management
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Long conversations consume context window tokens. Open Agents uses progressive context compaction to compress older messages while preserving critical information — decisions, errors, file states, and task progress.
 
 ### How It Works
@@ -1087,6 +1116,8 @@ This ensures the agent can resume coherently after compaction without re-reading
 
 ## Personality Core — SAC Framework Style Control
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The personality system controls how the agent communicates — from silent operator to teacher mode. It's based on the **SAC framework** ([arXiv:2506.20993](https://arxiv.org/abs/2506.20993)) which models personality along five behavioral intensity dimensions rather than binary trait toggles.
 
 ```bash
@@ -1135,6 +1166,8 @@ The personality system draws on:
 
 ## Emotion Engine — Affective State Modulation
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The agent stack includes a real-time emotion system that modulates behavior based on an appraisal-based affective model. Built on Russell's circumplex model of affect extended with the dominance axis from UDDETTS ADV space ([arXiv:2505.10599](https://arxiv.org/abs/2505.10599)), the engine maintains a continuous emotional state defined by three axes:
 
 - **Valence** (-1 to +1): displeasure ↔ pleasure
@@ -1197,6 +1230,8 @@ The emotion system is informed by peer-reviewed and preprint research:
 
 ## Voice Feedback (TTS)
 
+<div align="right"><a href="#top">back to top</a></div>
+
 ```bash
 /voice              # Toggle on/off (default: GLaDOS)
 /voice glados       # GLaDOS voice (ONNX, ~50MB)
@@ -1388,6 +1423,8 @@ The stochastic narration engine generates spoken descriptions of what the agent
 
 ## Listen Mode — Live Bidirectional Audio
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Listen mode enables real-time voice communication with the agent. Your microphone audio is captured, streamed through Whisper, and the transcription is injected directly into the input line — creating a hands-free coding workflow.
 
 Two transcription backends ensure broad platform support:
@@ -1424,6 +1461,8 @@ The `transcribe-cli` dependency auto-installs in the background on first use. On
 
 ## Vision & Desktop Automation (Moondream)
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Open Agents can see your screen, understand UI elements, and interact with desktop applications through natural language — powered by the Moondream vision language model running entirely locally.
 
 ### Desktop Awareness
@@ -1610,6 +1649,8 @@ Supports `apt` (Debian/Ubuntu), `dnf` (Fedora), `pacman` (Arch), and `brew` (mac
 
 ## Interactive TUI
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Launch without arguments to enter the interactive REPL:
 
 ```bash
@@ -1713,6 +1754,8 @@ The steering sub-agent uses the same model and backend as the main agent with `m
 
 ## Telegram Bridge — Sub-Agent Per Chat
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Connect the agent to a Telegram bot. Each incoming message spawns a dedicated sub-agent that handles the conversation independently — visible in the terminal waterfall alongside other agent activity.
 
 ```bash
@@ -1844,6 +1887,8 @@ The bridge automatically handles Telegram's rate limits (HTTP 429) with exponent
 
 ## x402 Payment Rails & Nexus P2P
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Agents can earn and spend USDC on Base mainnet through the native x402 protocol built into [open-agents-nexus@1.5.6](https://www.npmjs.com/package/open-agents-nexus).
 
 ### Wallet & Identity
@@ -1901,6 +1946,8 @@ nexus(action='budget_set', auto_approve_below='0.01')  # Auto-approve micropayme
 
 ## Sponsored Inference — Share Your GPU With the World
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Anyone running Open Agents can become an inference sponsor — sharing their local models (or forwarded cloud endpoints) with users worldwide through a secure, branded relay.
 
 ### For Sponsors: `/sponsor`
@@ -1965,6 +2012,8 @@ The tunnel fix uses debounced restarts with exponential cooldown (10s → 20s
 
 ## Dream Mode — Creative Idle Exploration
 
+<div align="right"><a href="#top">back to top</a></div>
+
 When you're not actively tasking the agent, Dream Mode lets it creatively explore your codebase and generate improvement proposals autonomously. The system models real human sleep architecture with four stages per cycle:
 
 | Stage | Name | What Happens |
@@ -2039,6 +2088,8 @@ If no GPU is detected, the REM stage falls back to the standard multi-agent crea
 
 ## Blessed Mode — Infinite Warm Loop
 
+<div align="right"><a href="#top">back to top</a></div>
+
 `/full-send-bless` activates an infinite warm loop that keeps model weights loaded in VRAM and the agent ready for instant response. The engine sends periodic keep-alive pings to the inference backend (every 2 minutes) to prevent Ollama's automatic model unloading.
 
 ```bash
@@ -2076,6 +2127,8 @@ Each DMN cycle runs a lightweight LLM agent (15 max turns, temperature 0.4) with
 
 ## Code Sandbox
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Execute code snippets in isolated environments without affecting your project:
 
 ```
@@ -2092,6 +2145,8 @@ Supports JavaScript, TypeScript, Python, and Bash. Two execution modes:
 
 ## Structured Data Tools
 
+<div align="right"><a href="#top">back to top</a></div>
+
 ### Generate structured files
 
 Create CSV, TSV, JSON, Markdown tables, and Excel-compatible files from data:
@@ -2118,6 +2173,8 @@ Detects binary formats (XLSX, PDF, DOCX) and suggests conversion tools.
 
 ## Multi-Provider Web Search
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Web search automatically selects the best available provider:
 
 | Provider | Trigger | Features |
@@ -2133,6 +2190,8 @@ export JINA_API_KEY=jina_...     # Enable Jina AI (optional)
 
 ## Task Templates
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Set a task type to get specialized system prompts, recommended tools, and output guidance:
 
 ```
@@ -2144,6 +2203,8 @@ Set a task type to get specialized system prompts, recommended tools, and output
 
 ## Human Expert Speed Ratio
 
+<div align="right"><a href="#top">back to top</a></div>
+
 The status bar displays a real-time `Exp: Nx` gauge estimating how fast the agent is working relative to a leading human expert performing equivalent tasks.
 
 ```
@@ -2183,6 +2244,8 @@ All 47 tools have calibrated baselines ranging from 3s (`task_stop`) to 180s (`c
 
 ## Cost Tracking & Session Metrics
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Real-time token cost estimation for cloud providers. The status bar shows running cost when using a paid endpoint.
 
 ```
@@ -2197,6 +2260,8 @@ Work evaluation uses five task-type-specific rubrics (code, document, analysis,
 
 ## Configuration
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Config priority: CLI flags > env vars > `~/.open-agents/config.json` > defaults.
 
 ```bash
@@ -2227,6 +2292,8 @@ Create `AGENTS.md`, `OA.md`, or `.open-agents.md` in your project root for agent
 
 ## Model Support
 
+<div align="right"><a href="#top">back to top</a></div>
+
 **Primary target**: Qwen3.5-122B-A10B via Ollama (MoE, 48GB+ VRAM)
 
 Any Ollama or OpenAI-compatible API model with tool calling works:
@@ -2239,6 +2306,8 @@ oa --backend-url http://10.0.0.5:11434 "refactor auth"
 
 ## Supported Inference Providers
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Open Agents auto-detects your provider from the endpoint URL and configures auth + health checks accordingly. All providers use standard `Authorization: Bearer <key>` authentication.
 
 | Provider | Endpoint URL | API Key | Notes |
@@ -2395,6 +2464,8 @@ No configuration needed — the cascade is built from your endpoint usage histor
 
 ## Evaluation Suite
 
+<div align="right"><a href="#top">back to top</a></div>
+
 46 evaluation tasks test the agent's autonomous capabilities across coding, web research, SDLC analysis, tool creation, multi-file reasoning, memory systems, and context engineering:
 
 ```bash
@@ -2510,6 +2581,8 @@ The PoT (Program-of-Thought) guidance achieves **100% code generation rate** —
 
 ## AIWG Integration
 
+<div align="right"><a href="#top">back to top</a></div>
+
 Open Agents integrates with [AIWG](https://aiwg.io) ([npm](https://www.npmjs.com/package/aiwg)) for AI-augmented software development:
 
 ```bash
@@ -2527,6 +2600,8 @@ oa "analyze this project's SDLC health and set up documentation"
 
 ## License
 
+<div align="right"><a href="#top">back to top</a></div>
+
 [Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)](https://creativecommons.org/licenses/by-nc/4.0/)
 
 Free for non-commercial use. For enterprise/commercial licensing, contact [zoomerconsulting.com](https://zoomerconsulting.com).

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "open-agents-ai",
-  "version": "0.185.70",
+  "version": "0.185.71",
   "description": "AI coding agent powered by open-source models (Ollama/vLLM) — interactive TUI with agentic tool-calling loop",
   "type": "module",
   "main": "./dist/index.js",