sciClaw

https://sciclaw.dev
Admin
sciClaw is an AI-powered research assistant for scientists that combines LLM capabilities with...
Tokens:16,970
Snippets:280
Trust Score:1.9
Update:6 months ago
Show doc for...
Tokens:
Raw
### Initialize sciClaw (App Recommended)

Source: https://sciclaw.dev/docs

Start the sciClaw application, which opens the dashboard and initiates the setup wizard for a guided experience.

```bash
# Recommended: opens dashboard + starts setup wizard
sciclaw app
```

--------------------------------

### Run Initial Setup Flow

Source: https://sciclaw.dev/docs

Executes the first-time setup for sciClaw, creating configuration files, workspace directories, and installing necessary components. Use the --yes flag for non-interactive installation.

```bash
sciclaw onboard
```

```bash
sciclaw onboard --yes
```

--------------------------------

### Install and Start sciClaw Gateway

Source: https://sciclaw.dev/docs

Use these commands to install the sciClaw service in the background and then start it. This is the recommended mode for daily use.

```bash
sciclaw service install
sciclaw service start
```

```bash
sciclaw gateway
```

--------------------------------

### Install and Start sciClaw Service

Source: https://sciclaw.dev/section-install.html

Installs the sciClaw service to run in the background (e.g., via launchd/systemd) and then starts it.

```bash
sciclaw service install && sciclaw service start
```

--------------------------------

### Initialize sciClaw Application

Source: https://sciclaw.dev/docs

Use `sciclaw app` to open the dashboard and start the setup wizard. For headless environments, `sciclaw onboard` provides a CLI alternative.

```bash
# Recommended: opens the dashboard and starts first-run setup
sciclaw app

# CLI alternative for headless/non-TUI environments
sciclaw onboard
```

--------------------------------

### Install Skills from Catalog

Source: https://sciclaw.dev/docs

Install additional skills from the skills catalog. The installer validates content before writing to disk, including size limits and provenance.

```bash
sciclaw skills install drpedapati/sciclaw-skills/<skill-name>
```

--------------------------------

### Install sciClaw

Source: https://sciclaw.dev/v0.1.56-announcement.md

Tap the custom Homebrew repository and install sciClaw using 'brew install'. This command sets up sciClaw on your system. After installation, run 'sciclaw app' to launch the application.

```bash
brew tap drpedapati/tap && brew install sciclaw
sciclaw app
```

--------------------------------

### Install sciClaw from Source

Source: https://sciclaw.dev/docs

Clones the sciClaw repository, installs dependencies, and builds the binary for your platform. Requires Go 1.25+ and make.

```bash
git clone https://github.com/drpedapati/sciclaw.git
cd sciclaw
make deps

# Build for your platform
make build

# Or build and install to ~/.local/bin
make install
```

--------------------------------

### Start Ollama service on Linux

Source: https://sciclaw.dev/docs

Command to start the Ollama service if it is installed but not running.

```bash
systemctl start ollama
```

--------------------------------

### Install or update ctxclaw

Source: https://sciclaw.dev/docs

Standalone installation instructions for the ctxclaw binary.

```bash
brew install drpedapati/tap/ctxclaw
ctxclaw version
```

--------------------------------

### Verify Homebrew Installation

Source: https://sciclaw.dev/docs

Confirm that Homebrew has been successfully installed by checking its version.

```bash
brew --version
```

--------------------------------

### Install TeX Live for PDF Rendering

Source: https://sciclaw.dev/docs

Installs necessary TeX Live packages for PDF rendering on Linux systems where TinyTeX auto-install is unavailable. Requires sudo privileges.

```bash
sudo apt-get update
sudo apt-get install -y texlive-xetex texlive-fonts-recommended texlive-latex-extra
```

--------------------------------

### Install sciClaw with Homebrew

Source: https://sciclaw.dev/section-install.html

Installs sciClaw and all companion tools using Homebrew. Pre-compiled binaries are used, so no build step is required.

```bash
brew tap drpedapati/tap && brew install sciclaw
```

--------------------------------

### Set up sciClaw Workspace

Source: https://sciclaw.dev/section-install.html

Creates the `~/sciclaw` folder with necessary files and configuration. An interactive wizard guides through authentication and optional tools.

```bash
sciclaw onboard
```

--------------------------------

### Set up sciClaw Workspace with Defaults

Source: https://sciclaw.dev/section-install.html

Skips the interactive wizard and accepts default settings for workspace setup.

```bash
sciclaw onboard --yes
```

--------------------------------

### Install a Skill

Source: https://sciclaw.dev/docs

Add a new skill to your sciClaw installation.

```bash
sciclaw skills install
```

--------------------------------

### Install Companion CLI Tools

Source: https://sciclaw.dev/docs

Install standalone CLI tools that provide native file format support for Word documents and PubMed searches. These should be installed alongside sciClaw for full functionality.

```bash
brew install drpedapati/tools/docx-review   # Word docs with tracked changes (Open XML SDK)
brew install drpedapati/tools/pubmed-cli     # PubMed search & citation graphs
```

--------------------------------

### Install Homebrew

Source: https://sciclaw.dev/docs

Install Homebrew package manager on macOS or Linux by running the provided script in your terminal.

```bash
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
```

--------------------------------

### List Installed Skills

Source: https://sciclaw.dev/docs

Display a list of all skills that are currently installed in the sciClaw workspace.

```bash
sciclaw skills list
```

--------------------------------

### Upgrade and Onboard Existing Installations

Source: https://sciclaw.dev/v0.1.54-announcement.md

Update an existing sciClaw installation and run diagnostic tools to ensure proper configuration.

```bash
brew upgrade sciclaw
sciclaw onboard    # refresh skills
sciclaw doctor     # verify
```

--------------------------------

### Check Quarto and PDF Toolchain

Source: https://sciclaw.dev/docs

Verifies the installation and configuration of Quarto and its associated PDF rendering tools.

```bash
quarto check
```

--------------------------------

### Verify System Components

Source: https://sciclaw.dev/docs

Commands to validate the installation, check component status, and perform a basic connectivity test.

```bash
# Check version
sciclaw --version

# Check all components: config, workspace, providers, IRL runtime
sciclaw status

# Verify skills are loaded
sciclaw skills list

# Quick smoke test
sciclaw agent -m "Hello, what version are you?"
```

--------------------------------

### Install Office Tool Repo Migration CLIs

Source: https://sciclaw.dev/docs

Install office companion tools using the new tap. These tools were transferred from henrybloomingdale/* to drpedapati/*.

```bash
brew tap drpedapati/tools
brew install drpedapati/tools/docx-review
brew install drpedapati/tools/pptx-review
brew install drpedapati/tools/xlsx-review
```

--------------------------------

### Manage Background Gateway Service

Source: https://sciclaw.dev/docs

Control the background sciClaw gateway service using subcommands like install, start, stop, restart, status, and logs.

```bash
sciclaw service <subcommand>
```

--------------------------------

### Per-Invocation Model Override Examples

Source: https://sciclaw.dev/docs

Examples showing how to use the '--model' flag to override the default model for a single command execution.

```bash
# Quick task with GPT-5.2 (config unchanged)
sciclaw agent --model gpt-5.2 -m "Summarize this CSV"

# Deep analysis with maximum GPT-5.2 effort
sciclaw agent --model gpt-5.2 --effort xhigh -m "Review this manuscript draft"

# Interactive session with a specific model
sciclaw agent --model gpt-5.2

```

--------------------------------

### Example: Execute a Full Analysis Protocol

Source: https://sciclaw.dev/index.html

Provide a protocol file to execute a multi-step research workflow, including data pulling, preprocessing, statistical analysis, and manuscript rendering.

```natural_language
Follow analysis-protocol.md. Pull the latest data from REDCap, run the preprocessing pipeline in Python, then the stats in R. Output a Quarto manuscript with inline results. Log everything.
```

--------------------------------

### Manage SciClaw Service Lifecycle

Source: https://sciclaw.dev/docs

Control the SciClaw gateway service, which runs in the background. Commands include installation, starting, stopping, status checks, and log retrieval.

```bash
# Install + start background gateway
sciclaw service install
sciclaw service start

# Lifecycle
sciclaw service status
sciclaw service restart
sciclaw service stop
sciclaw service logs --lines 200
sciclaw service uninstall
```

--------------------------------

### Set up Telegram Chat Channel

Source: https://sciclaw.dev/section-install.html

Guides through the process of creating a Telegram bot and linking it to sciClaw.

```bash
sciclaw channels setup telegram
```

--------------------------------

### Start sciClaw Gateway in Terminal

Source: https://sciclaw.dev/section-install.html

Starts the sciClaw gateway directly in the current terminal session.

```bash
sciclaw gateway
```

--------------------------------

### Connect AI Provider - OpenAI/Anthropic

Source: https://sciclaw.dev/docs

During setup, choose an AI provider. OpenAI uses device-code login without an API key. Anthropic requires an API key from their console.

```bash
 Setup: Authentication
╭──────────────────────────────────────────────────────╮
│ ✓ Configuration file created.                        │
│                                                       │
│ Choose your AI provider:
│                                                       │
│ [Enter]  Log in with OpenAI (recommended)             │
│ [a]      Use Anthropic API key                        │
│ [Esc]    Skip for now                                │
╰──────────────────────────────────────────────────────╯
                    Step 2 of 5
```

--------------------------------

### Migrate from OpenClaw

Source: https://sciclaw.dev/docs

Execute the migration process from an OpenClaw installation to sciClaw.

```bash
sciclaw migrate
```

--------------------------------

### Launch TUI Dashboard

Source: https://sciclaw.dev/docs

Command to start the visual terminal-based dashboard.

```bash
sciclaw app
```

--------------------------------

### Display Version Information

Source: https://sciclaw.dev/docs

Shows the currently installed version of the sciClaw CLI.

```bash
sciclaw --version
```

--------------------------------

### Example: Specify Statistical Method and Parameters

Source: https://sciclaw.dev/index.html

Define the exact statistical analysis, including correlation type, correction methods, and data source. This example uses R syntax.

```natural_language
Run Spearman correlations between P300 amplitude at Pz and ALSFRS-R decline in R. Correct for multiple comparisons with Benjamini-Hochberg at FDR 0.05. Use erp_data.csv.
```

--------------------------------

### Install phi-cleaner CLI

Source: https://sciclaw.dev/docs

Install the phi-cleaner CLI tool, a helper for clinical text de-identification, if needed. This is an optional bundled skill.

```bash
brew tap drpedapati/tools
brew install drpedapati/tools/phi-cleaner
```

--------------------------------

### Agent Interaction Examples

Source: https://sciclaw.dev/docs

Demonstrates various ways to interact with the sciClaw agent, including one-shot queries, interactive sessions, and model/effort overrides.

```bash
# One-shot query
sciclaw agent -m "Summarize recent ALS biomarker literature"

# Interactive session
sciclaw agent

# Override model for one invocation (config unchanged)
sciclaw agent --model gpt-5.2 -m "Quick question"

# Set reasoning effort for GPT-5.2 (quality/latency/cost control)
sciclaw agent --effort high -m "Analyze this complex dataset"

# Combine both overrides
sciclaw agent --model gpt-5.2 --effort xhigh -m "Deep analysis"

```

--------------------------------

### Upgrade sciClaw and Run Application

Source: https://sciclaw.dev/v0.1.56-announcement.md

Use 'brew upgrade' to update sciClaw to the latest version. Then, run 'sciclaw app' to start the application. This sequence ensures you are using the most recent features and fixes.

```bash
brew upgrade sciclaw
sciclaw app
```

--------------------------------

### Configure HOOKS.md Policy

Source: https://sciclaw.dev/docs

Define plain-language instructions for lifecycle events to guide agent capture behavior.

```markdown
## before_turn
- Capture project context needed for reproducibility.

## after_turn
- Summarize completed actions and unresolved risks.
- Record what should be updated in manuscript and logs.

## before_tool
- Record the intent of the tool call and key parameters (redacting secrets).

## after_tool
- Record tool outcomes and where artifacts were written.

## on_error
- Capture failure context, likely cause, and the next recovery step.
```

--------------------------------

### Configure Telegram Channel

Source: https://sciclaw.dev/docs

Setup and manual configuration for Telegram bot integration.

```bash
sciclaw channels setup telegram
```

```json
{
  "channels": {
    "telegram": {
      "enabled": true,
      "token": "YOUR_BOT_TOKEN",
      "allow_from": ["YOUR_USER_ID"]
    }
  }
}
```

--------------------------------

### PHI Mode Configuration Schema

Source: https://sciclaw.dev/docs

Example of the configuration structure used for local inference settings.

```json
{
  "agents": {
    "defaults": {
      "mode": "phi",
      "local_backend": "ollama",
      "local_model": "qwen3.5:4b",
      "local_preset": "balanced"
    }
  }
}
```

--------------------------------

### List Managed IRL Projects

Source: https://sciclaw.dev/docs

Get a structured JSON view of all managed projects for discoverability and context-switching between studies.

```bash
irl list --json
```

--------------------------------

### Example: Request Specific Output Format

Source: https://sciclaw.dev/index.html

Specify the desired output format, such as a Word document with tracked changes, for a drafted section.

```natural_language
Draft a methods section and give me a Word doc with tracked changes.
```

--------------------------------

### Start Interactive Agent Session

Source: https://sciclaw.dev/docs

Initiate an interactive chat session with the sciClaw agent for ongoing conversation and tasks.

```bash
sciclaw agent
```

--------------------------------

### Setup Discord Channel for sciClaw

Source: https://sciclaw.dev/docs

This command sets up the Discord channel for sciClaw. Remember to enable the MESSAGE CONTENT INTENT in the Discord Developer Portal.

```bash
sciclaw channels setup discord
```

--------------------------------

### Example: Summarize Data with Missing Values

Source: https://sciclaw.dev/index.html

Provide a file path to have sciClaw summarize demographic data and identify any missing information.

```natural_language
Summarize the demographics in participants.csv and flag any missing data.
```

--------------------------------

### Adopt Existing Directory into IRL Management

Source: https://sciclaw.dev/docs

Bring existing research directories under IRL management without needing to start over. This allows for retroactively organizing past work.

```bash
irl adopt ./messy-folder
```

--------------------------------

### Check sciClaw Status

Source: https://sciclaw.dev/docs

Retrieves the current status of the sciClaw installation, which can be useful before and after upgrades.

```bash
sciclaw status
```

--------------------------------

### Workspace Directory Structure

Source: https://sciclaw.dev/docs

The default workspace directory structure for sciClaw, including directories for sessions, memory, state, and installed skills.

```bash
~/sciclaw/
├── sessions/          # Conversation history
├── memory/            # Long-term memory (MEMORY.md)
├── state/             # Persistent state
├── cron/              # Scheduled jobs database
├── skills/            # Installed skills
├── hooks/             # Hook audit log (hook-events.jsonl)
├── irl/
│   └── commands/      # IRL command execution audit records
├── AGENTS.md          # Agent behavior guide
├── HEARTBEAT.md       # Periodic task prompts
├── HOOKS.md           # Hook policy (plain-language)
├── IDENTITY.md        # sciClaw identity
├── SOUL.md            # Agent values & guardrails
├── TOOLS.md           # Tool descriptions
└── USER.md            # User preferences

```

--------------------------------

### Manage SciClaw Skills

Source: https://sciclaw.dev/docs

Extend SciClaw's capabilities by installing and managing skills. Skills are directories containing specialized knowledge and workflows, defined by a `SKILL.md` file.

```bash
# List installed skills
sciclaw skills list

# Install a skill from a directory or URL
sciclaw skills install ./my-skill
```

--------------------------------

### Example: Quick Question about TDP-43

Source: https://sciclaw.dev/index.html

Initiate a research query by asking a concise question. sciClaw will search PubMed, retrieve relevant papers, and summarize key findings.

```natural_language
What do we know about TDP-43 in ALS?
```

--------------------------------

### Set Default AI Model

Source: https://sciclaw.dev/section-install.html

Sets the default AI model to be used by sciClaw. This example sets it to Anthropic's Claude Sonnet 4.6.

```bash
sciclaw models set anthropic/claude-sonnet-4.6
```

--------------------------------

### List Available Models and Providers

Source: https://sciclaw.dev/docs

Display all configured providers and the models they offer.

```bash
sciclaw models list
```

--------------------------------

### Initialize New IRL Project

Source: https://sciclaw.dev/docs

Create a new, auto-named, datestamped research project with consistent directory structure, plans, data, and manuscript scaffolding.

```bash
irl init "ERP correlation analysis"
```

--------------------------------

### Initialize PHI Mode

Source: https://sciclaw.dev/docs

Run these commands to set up the local Ollama backend and verify the current status.

```bash
# One-time setup: detect hardware, pull model, verify
sciclaw modes phi-setup

# Confirm it's active
sciclaw modes status
```

--------------------------------

### Update and Verify Configuration

Source: https://sciclaw.dev/docs

Commands to inspect current configuration settings and model provider status.

```bash
# View current config
sciclaw status

# Check model and provider details
sciclaw models status
```

--------------------------------

### Manage authentication via CLI

Source: https://sciclaw.dev/docs

Use these commands to log in, check status, or log out of authentication providers.

```bash
# OpenAI device-code login (ChatGPT Plus/Team/Enterprise)
sciclaw auth login --provider openai

# Anthropic auth (API key or supported oat-token login flow)
sciclaw auth login --provider anthropic

# Check status
sciclaw auth status

# Remove stored credential
sciclaw auth logout --provider openai
```

--------------------------------

### Configure and Manage Routing via CLI

Source: https://sciclaw.dev/docs

Use these commands to automate channel-to-workspace mappings and debug routing logic.

```bash
sciclaw routing add \
  --channel discord \
  --chat-id 1467290563052048476 \
  --workspace /absolute/path/to/project-a \
  --allow 8535331528 \
  --label project-a \
  --no-mention

sciclaw routing enable
sciclaw routing validate
sciclaw routing reload

# Debug why a user message did/didn't route
sciclaw routing explain \
  --channel discord \
  --chat-id 1467290563052048476 \
  --sender 8535331528 \
  --mention
```

--------------------------------

### Remove Legacy Homebrew Tap

Source: https://sciclaw.dev/section-install.html

Removes the deprecated legacy Homebrew tap for sciClaw if it is still installed.

```bash
brew untap drpedapati/sciclaw
```

--------------------------------

### Manage Inference Modes

Source: https://sciclaw.dev/docs

Commands to switch between cloud, local, and VM inference modes.

```bash
# Show current mode, backend, model, hardware
sciclaw modes status

# Switch to local inference
sciclaw modes set phi

# Switch back to cloud
sciclaw modes set cloud

# Switch to VM mode
sciclaw modes set vm

# Re-run setup (e.g., after hardware change)
sciclaw modes phi-setup
```

--------------------------------

### Log in to OpenAI Account

Source: https://sciclaw.dev/section-install.html

Initiates the login process for the OpenAI provider using a device code flow. No API key is needed.

```bash
sciclaw auth login --provider openai
```

--------------------------------

### Check System Configuration

Source: https://sciclaw.dev/docs

Verifies that all sciClaw components and dependencies are correctly configured and reachable. Provides explanations and suggested fixes for any detected issues.

```bash
sciclaw doctor
```

--------------------------------

### Run sciClaw with Docker

Source: https://sciclaw.dev/docs

Commands to clone, configure, and execute the sciClaw agent or gateway using Docker Compose.

```bash
# Clone and configure
git clone https://github.com/drpedapati/sciclaw.git
cd sciclaw
cp config/config.example.json config/config.json

# Start gateway mode
docker compose --profile gateway up -d

# Or run agent one-shot
docker compose run --rm sciclaw-agent -m "Hello"

# Check logs
docker compose logs -f sciclaw-gateway

# Stop
docker compose --profile gateway down
```

--------------------------------

### Verify IRL Availability

Source: https://sciclaw.dev/docs

Check the status of the IRL runtime integration.

```bash
sciclaw status
# Output includes: IRL Runtime: ✓ /opt/homebrew/bin/irl
```

--------------------------------

### Configure Web Server to Listen on All Interfaces

Source: https://sciclaw.dev/docs

Modify the systemd user service to allow the web server to accept connections from other machines on the network. This involves stopping the service, editing its configuration to bind to 0.0.0.0, and then reloading and restarting the service.

```bash
# Stop the default service
systemctl --user stop sciclaw-web

# Edit the service to listen on all interfaces
systemctl --user edit sciclaw-web --force

# Add this content:
[Service]
ExecStart=
ExecStart=/home/linuxbrew/.linuxbrew/bin/sciclaw web --listen 0.0.0.0:4142

# Reload and restart
systemctl --user daemon-reload
systemctl --user start sciclaw-web
```

--------------------------------

### Access Web UI on Localhost

Source: https://sciclaw.dev/docs

Open this URL in a browser when sciClaw is running on the same machine to access the web UI.

```bash
http://localhost:4142
```

--------------------------------

### Run SciClaw Deployment Diagnostics

Source: https://sciclaw.dev/docs

Verify the entire SciClaw deployment, including configuration, workspace, authentication, CLI tools, skills, gateway, and service health. Provides both human-readable and machine-readable output.

```bash
# Human-readable report
sciclaw doctor

# Machine-readable output
sciclaw doctor --json
```

--------------------------------

### Inspect session prompt tokens as JSON

Source: https://sciclaw.dev/docs

Retrieve machine-readable token breakdown for a session.

```bash
sciclaw prompt inspect --session <key> --json
```

--------------------------------

### Re-install a Third-Party Skill

Source: https://sciclaw.dev/docs

Updates a specific third-party skill by re-installing it from its GitHub repository. Replace 'username/repo/skill-name' with the actual skill identifier.

```bash
sciclaw skills install username/repo/skill-name
```

--------------------------------

### Authenticate with a Provider

Source: https://sciclaw.dev/docs

Log in and authenticate with a sciClaw service provider to enable access to their models and services.

```bash
sciclaw auth login
```

--------------------------------

### Configure sciClaw settings

Source: https://sciclaw.dev/docs

The main configuration file located at ~/.picoclaw/config.json controls agent defaults, tool settings, and heartbeat behavior.

```json
{
  "agents": {
    "defaults": {
      "workspace": "~/sciclaw",
      "model": "gpt-5.2",
      "reasoning_effort": "high",
      "max_tokens": 8192,
      "temperature": 0.7,
      "max_tool_iterations": 0
    }
  },
  "providers": { ... },
  "channels": { ... },
  "tools": {
    "web": {
      "brave": {
        "enabled": true,
        "api_key": "YOUR_BRAVE_API_KEY",
        "max_results": 5
      },
      "duckduckgo": {
        "enabled": true,
        "max_results": 5
      }
    },
    "pubmed": {
      "api_key": "YOUR_NCBI_API_KEY"
    }
  },
  "heartbeat": {
    "enabled": true,
    "interval": 30
  }
}
```

--------------------------------

### Inspect session prompt tokens

Source: https://sciclaw.dev/docs

Analyze token consumption for a specific session.

```bash
sciclaw prompt inspect --session <session-key>
```

--------------------------------

### Re-install Built-in Skills

Source: https://sciclaw.dev/docs

Re-installs all baseline scientific skills to their default locations, overwriting any existing versions. This is useful for refreshing skills to their latest built-in versions.

```bash
sciclaw skills install-builtin
```

--------------------------------

### Show Current Model and Provider Status

Source: https://sciclaw.dev/docs

Display the current default model, its resolved provider, authentication method, and configured effort level.

```bash
sciclaw models status
```

--------------------------------

### AI Provider Connection Confirmation

Source: https://sciclaw.dev/docs

After connecting, a smoke test confirms the AI provider is successfully connected and responsive.

```bash
╭──────────────────────────────────────────────────────╮
│ ✓ Configuration file created.                        │
│ ✓ Connected!                                         │
│                                                       │
│ Your AI responded:
│ │ Hello! I'm ready to help with your research.        │
│                                                       │
│ Press Enter to continue.                              │
╰──────────────────────────────────────────────────────╯
```

--------------------------------

### Update sciClaw from Source

Source: https://sciclaw.dev/docs

Updates the sciClaw source code, reinstalls dependencies, and rebuilds/reinstalls the binary to ~/.local/bin. Requires manual updates for the IRL component.

```bash
cd sciclaw
git pull
make deps
make install
```

--------------------------------

### Optimize prompt context pipeline

Source: https://sciclaw.dev/docs

Pipe session data through ctxclaw to reduce prompt size.

```bash
sciclaw prompt export --session <key> | ctxclaw optimize | sciclaw prompt preview
```

--------------------------------

### System Maintenance Commands

Source: https://sciclaw.dev/docs

Commands for verifying deployment health, managing background services, and handling authentication.

```APIDOC
## CLI: System & Auth

### Description
Utilities for maintaining the sciClaw environment, including health checks and credential management.

### Commands
- **sciclaw doctor**: Verify deployment (config, tools, skills, auth, gateway).
- **sciclaw doctor --fix**: Auto-fix issues like missing baseline skills.
- **sciclaw auth login**: Authenticate with a provider.
- **sciclaw auth status**: Show stored credentials.
- **sciclaw auth logout**: Remove stored credentials.
- **sciclaw service <subcommand>**: Manage background gateway service (start/stop/restart/status/logs).
```

--------------------------------

### Update sciClaw and Check Doctor

Source: https://sciclaw.dev/docs

Use Homebrew to upgrade sciClaw to the latest version and then run `sciclaw doctor` to check for any missing companion tools or configuration issues.

```bash
brew upgrade sciclaw
sciclaw doctor
```

--------------------------------

### Check PHI mode status

Source: https://sciclaw.dev/docs

Verify the current AI mode configuration.

```bash
sciclaw modes phi-status
```

--------------------------------

### Configure provider API keys

Source: https://sciclaw.dev/docs

Directly add API keys to the providers section of the configuration file for services not using OAuth.

```json
"providers": {
  "openrouter": { "api_key": "sk-or-v1-..." },
  "anthropic":  { "api_key": "sk-ant-..." },
  "groq":       { "api_key": "gsk_..." }
}
```

--------------------------------

### Configure Heartbeat Settings

Source: https://sciclaw.dev/docs

JSON configuration for enabling and setting the interval of the heartbeat system.

```json
{
  "heartbeat": {
    "enabled": false,
    "interval": 30
  }
}
```

--------------------------------

### Interact with sciClaw Agent via CLI

Source: https://sciclaw.dev/section-install.html

Allows direct interaction with the sciClaw agent from the command line interface by providing a question.

```bash
sciclaw agent -m "your question"
```

--------------------------------

### Show Stored Authentication Credentials

Source: https://sciclaw.dev/docs

View the currently stored authentication credentials for sciClaw service providers.

```bash
sciclaw auth status
```

--------------------------------

### Manage sciClaw Service

Source: https://sciclaw.dev/docs

Commands to run the sciClaw gateway in background or foreground modes.

```bash
# Always-on background mode (recommended)
sciclaw service install
sciclaw service start

# Or foreground mode
sciclaw gateway
```

--------------------------------

### Configure Collaborative Routing

Source: https://sciclaw.dev/v0.1.54-announcement.md

Map a specific communication channel to a workspace folder to restrict agent access to project-specific files.

```bash
sciclaw routing add \
  --channel discord \
  --chat-id 123456 \
  --workspace /data/project-a \
  --allow alice,bob
```

--------------------------------

### Set Default Model and Reasoning Effort in Config

Source: https://sciclaw.dev/docs

Configure the default AI model and reasoning effort directly in the `~/.picoclaw/config.json` file. This file is also updated by the `sciclaw models set` command.

```json
{
  "agents": {
    "defaults": {
      "model": "gpt-5.2",
      "reasoning_effort": "high"
    }
  }
}
```

--------------------------------

### Configure Azure OpenAI Service

Source: https://sciclaw.dev/docs

Azure requires a specific endpoint URL and model routing configuration in the config file.

```json
"azure": {
  "api_key": "your-azure-azure-api-key",
  "api_base": "https://YOUR-RESOURCE.openai.azure.com/openai/deployments/YOUR-DEPLOYMENT/chat/completions?api-version=2024-12-01-preview"
}
```

```json
"agents": {
  "defaults": {
    "model": "azure/your-deployment-name",
    "provider": "azure"
  }
}
```

--------------------------------

### Interact with sciClaw Agent

Source: https://sciclaw.dev/docs

Use natural language commands to manage research projects through the sciClaw agent.

```bash
sciclaw agent -m "Create a new project for ERP correlation analysis"

sciclaw agent -m "What projects do I have?"

sciclaw agent -m "Adopt my old als-biomarker folder as a managed project"
```

--------------------------------

### Manage Scheduled Cron Tasks

Source: https://sciclaw.dev/docs

Commands to list and add recurring tasks using the sciClaw CLI.

```bash
sciclaw cron list
sciclaw cron add "Check email" --every 2h
```

--------------------------------

### View Audit Log Events

Source: https://sciclaw.dev/docs

Use standard command-line tools to inspect the JSONL audit log for recent hook activity.

```bash
# View recent hook events
tail -5 ~/sciclaw/hooks/hook-events.jsonl | jq .
```

--------------------------------

### Check Gateway and Web Server Status

Source: https://sciclaw.dev/docs

Use these commands to check the independent status of the sciClaw gateway and web server services.

```bash
sciclaw service gateway status
```

```bash
sciclaw service web status
```

--------------------------------

### Agent Interaction Commands

Source: https://sciclaw.dev/docs

Commands for interacting with the sciClaw agent, including one-shot queries and interactive chat sessions.

```APIDOC
## CLI: sciclaw agent

### Description
Interact with the sciClaw agent. Supports one-shot queries via flags or full interactive chat sessions.

### Parameters
#### Options
- **-m** (string) - Optional - Send a one-shot message to the agent.
- **--model** (string) - Optional - Override the default model for this specific invocation.
- **--effort** (string) - Optional - Set the reasoning effort level (e.g., high, xhigh) for the invocation.

### Examples
```bash
# One-shot query
sciclaw agent -m "Summarize recent ALS biomarker literature"

# Interactive session
sciclaw agent

# Override model and effort
sciclaw agent --model gpt-5.2 --effort xhigh -m "Deep analysis"
```
```

--------------------------------

### Write Research Abstract Request

Source: https://sciclaw.dev/docs

Submit a request to search PubMed for specific research papers and generate a structured abstract, saving it to a file.

```bash
> Search PubMed for papers on TDP-43 proteinopathy in ALS
  published in the last 5 years. Then write a 250-word
  structured abstract summarizing the key findings and save
  it as abstract.docx in my workspace.
```

--------------------------------

### Model Management Commands

Source: https://sciclaw.dev/docs

Commands for inspecting, listing, and persistently updating the AI models used by sciClaw.

```APIDOC
## CLI: sciclaw models

### Description
Manage the AI models and providers configured for the sciClaw agent.

### Subcommands
- **list**: Show current model and configured providers.
- **set <model>**: Persistently change the default model.
- **effort <level>**: Persistently change the reasoning effort.
- **status**: Show model, provider, auth, and effort status.

### Examples
```bash
# List available models
sciclaw models list

# Set default model
sciclaw models set gpt-5.2

# Check status
sciclaw models status
```
```

--------------------------------

### List Scheduled Cron Jobs

Source: https://sciclaw.dev/docs

View all scheduled jobs that are configured to run via cron in sciClaw.

```bash
sciclaw cron list
```

--------------------------------

### Add a Scheduled Cron Job

Source: https://sciclaw.dev/docs

Schedule a new job to run automatically at specified intervals.

```bash
sciclaw cron add
```

--------------------------------

### Combine Model and Effort Overrides

Source: https://sciclaw.dev/docs

Use both '--model' and '--effort' flags simultaneously to specify a particular model and reasoning level for a single command.

```bash
sciclaw agent --model gpt-5.2 --effort xhigh -m "Deep analysis"
```

--------------------------------

### Auto-Fix sciClaw Deployment Issues

Source: https://sciclaw.dev/docs

Attempt to automatically fix common deployment issues, such as syncing baseline skills and removing legacy names.

```bash
sciclaw doctor --fix
```

--------------------------------

### Define HEARTBEAT.md Tasks

Source: https://sciclaw.dev/docs

Specify periodic tasks for the agent to execute in a markdown file.

```markdown
# Periodic Tasks

## Quick Tasks (respond directly)
- Report current time

## Long Tasks (use spawn for async)
- Search the web for AI news and summarize
- Check email and report important messages
```
Tokens:
Raw