Settings

Validation warnings (settings were saved)

General

Agent Backend

is not installed.

This backend requires an external tool:

Documentation

Also requires an external CLI:

API Key Required

Provider

Uses Claude Code CLI with Anthropic API. No API key needed if logged into Claude Code. Runs via Ollama's Anthropic Messages API compatibility (Claude Code v0.14.0+). Use any model on OpenRouter with full CLI tool access via their Anthropic-compatible API. Connect to any Anthropic-compatible endpoint (LiteLLM, vLLM, etc.). Route through LiteLLM proxy for HuggingFace, vLLM, Together AI, Groq, and 100+ providers.

Ollama Host

Ollama Model Run ollama list to see available models.

OpenRouter API Key Get your key at openrouter.ai/keys

Model Browse models at openrouter.ai/models

Base URL

API Key

Model

LiteLLM Proxy URL

LiteLLM API Key

Model Use provider/model format (e.g. huggingface/meta-llama/Llama-3-70b)

Max Output Tokens Set to match your model's limit (e.g. 8192 for DeepSeek). 0 = provider default.

Model Override Leave empty to let Claude Code auto-select the best model

Max Tool Turns Max agent tool-use loops per query (default: 25)

Provider

Uses Ollama's OpenAI-compatible API. No API key needed. Use any model on OpenRouter. Free and paid models available. Connect to any OpenAI-compatible endpoint (NVIDIA NIM, vLLM, etc.). Route through LiteLLM proxy for HuggingFace, vLLM, Together AI, Groq, and 100+ providers.

Ollama Host

Ollama Model

OpenRouter API Key Get your key at openrouter.ai/keys

Model Browse models at openrouter.ai/models

Base URL

API Key

Model

LiteLLM Proxy URL

LiteLLM API Key

Model Use provider/model format (e.g. huggingface/meta-llama/Llama-3-70b)

Max Output Tokens Set to match your model's limit (e.g. 8192 for DeepSeek). 0 = provider default.

Model

Max Turns

Requires Google ADK. Install with: pip install 'pocketpaw[google-adk]'

Provider

Route through LiteLLM proxy. Use provider/model format (e.g. openai/gpt-4o).

LiteLLM Proxy URL

LiteLLM API Key

Model Use provider/model format (e.g. huggingface/meta-llama/Llama-3-70b)

Max Output Tokens Set to match your model's limit (e.g. 8192 for DeepSeek). 0 = provider default.

Model

Max Turns

Requires Codex CLI installed. Install with: npm install -g @openai/codex

Model

Max Turns

Requires copilot CLI + pip install github-copilot-sdk

Provider

Model

Max Turns

Requires pip install deepagents. Uses LangChain's init_chat_model for multi-provider support.

Model Format: provider:model. Examples: anthropic:claude-sonnet-4-6, openai:gpt-4o, ollama:llama3.2, google_genai:gemini-2.0-flash, litellm:anthropic/claude-sonnet-4-6

LiteLLM Proxy URL

LiteLLM API Key

Use provider/model format after litellm: (e.g. litellm:huggingface/meta-llama/Llama-3-70b)

Ollama Host

OpenRouter API Key

Max Turns

Requires OpenCode server running. Start with: opencode --server

Server URL

Model

Bypass Permission Prompts Warning: Agent will execute actions without asking for approval.

API Keys

Keys are encrypted at rest. Only keys needed for your current backend and provider are shown.

Required for Current Backend

Anthropic API Key

Get Key Saved

Required for Claude SDK backend — get a key at console.anthropic.com

OpenAI API Key

Get Key Saved

Google API Key

Get Key Saved

Used for Gemini LLM and image generation. Free at AI Studio.

Tools & Services

Search

Tavily API Key

Get Key Saved

Brave Search API Key

Get Key Saved

Parallel AI API Key Saved

Used by both web search (parallel provider) and URL extract

Voice & Media

ElevenLabs API Key

Get Key Saved

Sarvam AI API Key

Get Key Saved

Indian language TTS (Bulbul), STT (Saaras), OCR (Vision), Translation (Mayura)

Google OAuth

Client ID

Get Key Saved

Client Secret Saved

Authorize Services

Gmail Calendar Drive Docs

Spotify

Client ID

Get Key Saved

Client Secret Saved

Authorize Spotify

Behavior & Safety

Tool Profile

Plan Mode Require approval before executing tools

Plan Mode Tools Comma-separated list of tools that require approval

Smart Model Routing Automatically select model based on task complexity (not recommended with Claude SDK — it has its own routing)

Simple Tier

Moderate Tier

Complex Tier

Safety

Injection Scanner Scan inbound messages for prompt injection attacks

LLM Deep Scan Use a secondary LLM to analyze suspicious content (requires API key)

PII Protection Detect and mask sensitive data (SSN, email, phone, credit card) before storing

Action:

Scan Memory Mask PII before writing to session history and long-term memory

Scan Audit Log Mask PII in audit log entries

Scan Log Output Scrub PII patterns from application log output

Memory

Backend

Auto-Learn Automatically extract facts from every conversation

Mem0 LLM Provider

LLM used for fact extraction and memory consolidation

Mem0 LLM Model

Embedder Provider

Embedder Model

Vector Store

Ollama Base URL

Search & Services

Web Search Provider

URL Extract Provider

Voice

TTS Provider

TTS Voice

Sarvam TTS Language

STT Provider

STT Model

OCR Provider

System

Web Server

Host

Port

Set host to 0.0.0.0 to allow access from other devices on your network

Click Restart Server after changing host or port to apply

Self-Audit Daemon Run automated security and health checks on a schedule

Schedule (cron) Cron expression (default: 3 AM daily)

Security Audit

Run 7 security checks on your configuration

/ checks passed

Self-Audit Reports

Health and security check history

No reports yet. Click "Run Now" or enable the daemon.

Report: — / passed

A2A Protocol Expose PocketPaw as a remote A2A agent and delegate tasks to external agents

Agent Name Name advertised in the Agent Card to other agents

Agent Description

Task Timeout (seconds) Maximum time to wait for an agent response before failing the task

Trusted Agents One URL per line. Only these agents can be used for outbound delegation (SSRF protection)

Soul Protocol

Persistent AI identity with psychology-informed memory, OCEAN personality, and emotional state.

Enable Soul Protocol Give the agent persistent identity, memory, and emotion

Soul Name

Archetype e.g. "The Coding Expert", "The Compassionate Creator"

Custom Persona

Auto-Save Interval (seconds) How often to auto-save soul state (0 = disabled, default 300 = 5 minutes)

Import / Export

Import

Import a .soul, .yaml, or .json file. Export saves the current soul to disk.