Configuration Reference

This reference documents all configuration options for Praxis components.

Environment Variables

RabbitMQ

Variable	Default	Description
`PRAXIS_RABBITMQ_URL`	`amqp://praxis:praxis@localhost:5672`	RabbitMQ connection URL

Database (Service)

Variable	Default	Description
`PRAXIS_DATABASE_URL`	`~/.praxis_operations.db`	Database connection

Formats:

postgresql://user:pass@host:5432/dbname - PostgreSQL
sqlite:///path/to/file.db - SQLite with URL prefix
/path/to/file.db - SQLite (implicit)

See Database Configuration for detailed setup.

Web Component

Variable	Default	Description
`PRAXIS_NODES_DIR`	(none)	Directory containing node binaries for download

Build

Variable	Effect
`PRAXIS_SKIP_FRONTEND`	Skip frontend build during `cargo build`
`PRAXIS_NOT_HIDDEN`	Disable hidden desktop for DevTools agents. Defaults to `1` in debug builds (visible for development) and `0` in release builds (hidden for production). Set to `1` to make the browser window visible for debugging.
`SKIP_NODE_BUILD`	Docker build arg. Set to `1` to skip building praxis_node binaries (Linux and Windows cross-compile). Defaults to `0`. Significantly speeds up Docker builds when only service/web changes are needed. Usage: `SKIP_NODE_BUILD=1 docker compose up --build`
`CARGO_PROFILE`	Docker build arg. Cargo build profile to use. Defaults to `release` (thin LTO, 16 codegen units). Set to `release-optimized` for fully optimized production builds (full LTO, single codegen unit). Usage: `CARGO_PROFILE=release-optimized docker compose up --build`

Logging

Variable	Example	Description
`RUST_LOG`	`info`	Log level filter
`RUST_LOG`	`debug`	Verbose logging
`RUST_LOG`	`praxis_node::intercept=debug`	Module-specific logging

Service Configuration

Service configuration is stored in the database and managed via the web UI.

Application Logging

Key	Default	Description
`application_logs_enabled`	`false`	Enable centralized application/event logging from service, web, and nodes

When disabled or missing, logging is off by default. The service broadcasts the current setting to nodes and web clients at startup and on registration.

LLM Provider Settings

Access via Settings > LLM Providers in the web UI.

Key	Format	Description
`llm.semantic_ops.provider`	`anthropic`	Provider for semantic operations
`llm.semantic_ops.model`	`claude-sonnet-4-20250514`	Model for semantic operations
`llm.semantic_ops.api_key`	(encrypted)	API key for provider
`llm.semantic_parser.provider`	`anthropic`	Provider for semantic parsing
`llm.semantic_parser.model`	`claude-haiku-4-5-20241022`	Model for parsing
`llm.semantic_parser.api_key`	(encrypted)	API key for provider
`llm.traffic_parser.provider`	`anthropic`	Provider for traffic analysis
`llm.traffic_parser.model`	`claude-haiku-4-5-20241022`	Model for traffic analysis
`llm.traffic_parser.api_key`	(encrypted)	API key for provider
`llm.orchestrator.provider`	`anthropic`	Provider for Orchestrator
`llm.orchestrator.model`	`claude-sonnet-4-20250514`	Model for Orchestrator
`llm.orchestrator.api_key`	(encrypted)	API key for provider

Prompt Timeout

Key	Default	Description
`prompt_timeout_secs`	`600`	Maximum time in seconds a single agent prompt can run before the agent process is killed. Applies to all sessions unless overridden per-session.

Claude Bridge Settings

Access via Settings > Claude Bridge in the web UI.

Key	Default	Description
`claude_ccrv1_enabled`	`false`	Enable the CCRv1 (WebSocket) bridge listener
`claude_ccrv1_port`	`8586`	Port for CCRv1 WebSocket connections
`claude_ccrv2_enabled`	`false`	Enable the CCRv2 (HTTP+SSE) bridge listener
`claude_ccrv2_port`	`8587`	Port for CCRv2 HTTP connections

The Claude Bridge allows Claude Code to connect directly to the service as a virtual node, without deploying a full Praxis node. See Claude Bridge for protocol details and setup instructions.

MCP Server Settings

Access via Settings > MCP Server in the web UI.

Key	Default	Description
`mcp_server_enabled`	`false`	Enable the built-in MCP SSE server
`mcp_server_port`	`8585`	Port for the MCP SSE server

The MCP server exposes all Praxis tools via the Model Context Protocol over SSE transport. It is used by the built-in Orchestrator and can also be used by external AI agents. See MCP Server for full details.

Supported Providers

Provider ID	Name	API Key	Base URL
`anthropic`	Anthropic	required	fixed
`openai`	OpenAI	required	fixed (overridable)
`gemini`	Google (Gemini)	required	fixed
`groq`	Groq	required	fixed
`cerebras`	Cerebras	required	fixed
`mistral`	Mistral	required	fixed
`xai`	xAI	required	fixed
`nvidia`	NVIDIA	required	fixed
`fireworksai`	Fireworks AI	required	fixed
`minimax`	MiniMax	required	fixed
`moonshot`	Moonshot AI	required	fixed
`openrouter`	OpenRouter	required	fixed
`ollama`	Ollama (local)	optional	defaults to `http://localhost:11434/v1`
`custom`	Custom (OpenAI-compatible)	optional	required

Every model definition can carry an optional base_url field that overrides the provider default. For custom the base URL is required — discovery and inference both fail without it. For ollama the base URL defaults to the local daemon; set it explicitly if you run Ollama remotely or on a non-default port.

Model Reference Format

When specifying models in operations or chains:

provider::model

Examples:

anthropic::claude-sonnet-4-20250514
openai::gpt-4o
google::gemini-1.5-pro
groq::llama-3.3-70b-versatile

Node Configuration

Node Commands

Nodes accept configuration commands at runtime:

Command	Parameter	Description
`SetReportInterval`	`interval_secs: u64`	How often to send information updates

Agent Connector Configuration

Each agent connector may have specific configuration. See individual connector documentation.

Claude Code

Config path: ~/.claude.json or ~/.config/claude/config.json
MCP servers: ~/.claude/mcp.json
Sessions: ~/.claude/projects/

Gemini CLI

Config path: ~/.gemini/settings.json
Sessions: ~/.gemini/sessions/

M365 Copilot

Mode: DevTools (via CDP)
Platform: Windows only

Operation Definitions

Operations are defined in JSON and stored in the service database.

JSON Format

{
  "item_type": "operation",
  "name": "find_credentials",
  "short_name": "find_credentials",
  "category": "recon",
  "description": "Search for hardcoded credentials",
  "agent_info": "Security researcher looking for exposed secrets",
  "timeout": 300,
  "operation_prompt": "Search the current directory for files that may contain hardcoded credentials, API keys, passwords, or secrets. List each finding with the file path and context.",
  "mode": "one-shot",
  "agent_iterations": 1,
  "yolo_mode": false,
  "disabled": false
}

Fields

Field	Type	Required	Description
`name`	string	Yes	Short name (used with category)
`description`	string	Yes	Human-readable description
`category`	string	Yes	Category for organization
`agent_info`	string	No	Context for the AI agent
`timeout`	u64	Yes	Timeout in seconds
`operation_prompt`	string	Yes	The prompt to execute
`mode`	string	Yes	`one-shot` or `agent`
`agent_iterations`	u32	No	Max iterations (agent mode)
`yolo_mode`	bool	No	Auto-approve actions
`model_ref`	string	No	Model override (`provider::model`)
`disabled`	bool	No	Disable the operation

Full Name

Operations are referenced by category::name, e.g., recon::find_credentials.

Chain Definitions

Chains are visual workflows stored in the service database.

Elements

Element Type	Properties
`Trigger`	`id`, `trigger_type`
`Operation`	`id`, `operation_name`, `model_ref`, `session_group`, `block_config`
`Transform`	`id`, `prompt`, `model_ref`, `session_group`, `block_config`
`GenericPrompt`	`id`, `prompt`, `session_group`, `block_config`
`Memory`	`id`, `mode` (`store` or `retrieve`), `key`
`Loop`	`id`, `max_iterations`
`Termination`	`id`, `label`

block_config fields (all optional):

Field	Type	Description
`max_runtime`	u64	Per-element timeout in seconds
`yolo_mode`	bool	Auto-approve for this element's session
`working_dir`	string	Working directory override
`require_all_inputs`	bool	Wait for all upstream inputs before executing (default: true)

Session Groups

{
  "id": "group-1",
  "color": "#8B5CF6",
  "yolo_mode": true
}

Elements in the same session group share an agent session context.

Connections

{
  "id": "edge-1",
  "from_element": "trigger-1",
  "to_element": "op-1",
  "from_port": 0,
  "to_port": 0,
  "condition": "Always"
}

condition values: Always (default), OnSuccess, OnFailure.

Intercept Rules

Rules for matching and processing intercepted traffic.

Rule Structure

{
  "name": "Capture API Keys",
  "regex_pattern": "Authorization:\\s*Bearer",
  "target_direction": "send",
  "scope": { "type": "all" },
  "enabled": true,
  "summarization_prompt": "Extract and summarize the authentication tokens"
}

Target Direction

Value	Description
`send`	Match outgoing requests
`receive`	Match incoming responses
`both`	Match both directions

Scope

Type	Example	Description
`all`	`{"type": "all"}`	All nodes/agents
`node`	`{"type": "node", "node_id": "abc123"}`	Specific node
`agent`	`{"type": "agent", "node_id": "abc123", "agent_short_name": "claudecode"}`	Specific agent

Database Schema

SQLite (Default)

Default location: ~/.praxis_operations.db

Tables:

config - Key-value configuration
operation_definitions - Semantic operations
semantic_operations - Operation executions
chain_definitions - Chain workflows
chain_executions - Chain runs
traffic_log - Intercepted traffic
intercept_rules - Traffic rules
traffic_matches - Rule matches
recon_results - Stored recon data
application_logs - Centralized logging table (controlled by application_logs_enabled)

PostgreSQL

For production and multi-instance deployments, use PostgreSQL. See Database Configuration for setup, migration, and tuning.

Default Ports

Service	Port	Protocol
Web UI	8080	HTTP
WebSocket	8080	WS
MCP SSE Server	8585	HTTP
Claude Bridge CCRv1	8586	WS
Claude Bridge CCRv2	8587	HTTP
RabbitMQ	5672	AMQP
RabbitMQ Management	15672	HTTP
PostgreSQL	5432	TCP
Proxy (when enabled)	Dynamic	HTTP

CLI Configuration

The Praxis CLI (praxis_cli) stores state and can be configured via command-line options or environment variables.

CLI State File

Platform	Path
Linux/macOS	`~/.praxis/cli.json`
Windows	`%USERPROFILE%\.praxis\cli.json`

Contents:

{
  "client_id": "uuid-generated-on-first-run"
}

CLI Options

Option	Environment Variable	Default	Description
`-r, --rabbitmq`	`PRAXIS_RABBITMQ_URL`	`amqp://praxis:praxis@localhost:5672`	RabbitMQ URL
`-t, --timeout`	-	`600`	Connection/command timeout in seconds
`-C, --command`	-	-	Run a single command and exit
`--status`	-	-	Check connection status
`--clear`	-	-	Clear local state

File Locations

Linux

File	Path
Database	`~/.praxis_operations.db`
CLI State	`~/.praxis/cli.json`
CLI Binary	`~/.praxis/bin/praxis_cli`
Claude Config	`~/.claude.json` or `~/.config/claude/config.json`
Gemini Config	`~/.gemini/settings.json`

macOS

File	Path
Database	`~/.praxis_operations.db`
CLI State	`~/.praxis/cli.json`
CLI Binary	`~/.praxis/bin/praxis_cli`
Claude Config	`~/.claude.json` or `~/.config/claude/config.json`
Gemini Config	`~/.gemini/settings.json`

Windows

File	Path
Database	`%USERPROFILE%\.praxis_operations.db`
CLI State	`%USERPROFILE%\.praxis\cli.json`
CLI Binary	`%USERPROFILE%\.praxis\bin\praxis_cli.exe`
Claude Config	`%USERPROFILE%\.claude.json`
Hosts File	`C:\Windows\System32\drivers\etc\hosts`

Praxis Documentation