### Geneclaw Quick Start Commands

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/marketing/twitter-thread.md

These commands provide a quick way to clone the Geneclaw repository, install dependencies, and run initial setup and evolution commands. No LLM key is required to use the heuristic-only mode.

```bash
git clone github.com/Clawland-AI/Geneclaw
pip install -e ".[dev,dashboard]"
nanobot onboard
nanobot geneclaw doctor
nanobot geneclaw evolve --dry-run
```

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

### Quick Start Examples for Summarize Skill

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/summarize/SKILL.md

Demonstrates basic usage for summarizing URLs, local files, and YouTube links using different models.

```bash
summarize "https://example.com" --model google/gemini-3-flash-preview
summarize "/path/to/file.pdf" --model google/gemini-3-flash-preview
summarize "https://youtu.be/dQw4w9WgXcQ" --youtube auto
```

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

### Install Geneclaw and nanobot

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/quickstart/Geneclaw-Runbook.md

Clone the Geneclaw repository and install it along with its development dependencies. This also includes the initial nanobot setup to create configuration files and a workspace.

```bash
git clone https://github.com/Clawland-AI/Geneclaw.git
cd Geneclaw
pip install -e ".[dev]"

nanobot onboard
```

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

### Install Dashboard Dependencies and Launch

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/dashboard-runbook.md

Installs the necessary dependencies for the dashboard and then launches the Streamlit application. The dashboard will be accessible at http://localhost:8501 by default.

```bash
pip install -e ".[dashboard]"
nanobot geneclaw dashboard
```

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

### Summarize Skill Configuration Example

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/summarize/SKILL.md

Shows an example of the optional configuration file for the Summarize skill, specifying a model.

```json
{
  "model": "openai/gpt-5.2"
}
```

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

### Initialize Nanobot Onboarding

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Run this command to start the onboarding process for Nanobot.

```bash
nanobot onboard
```

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

### Install Geneclaw from Source

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Use this command to clone the repository and install Geneclaw with development dependencies.

```bash
git clone https://github.com/Clawland-AI/Geneclaw.git
cd Geneclaw
pip install -e ".[dev]"
```

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

### Install a Skill using ClawHub

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/clawhub/SKILL.md

Installs a specified skill into the nanobot workspace. Always include the `--workdir` flag to ensure skills are placed in the correct directory (`~/.nanobot/workspace/skills/`). Replace `<slug>` with the actual skill name.

```bash
npx --yes clawhub@latest install <slug> --workdir ~/.nanobot/workspace
```

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

### List Installed Skills

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/clawhub/SKILL.md

Displays a list of all skills currently installed in the specified nanobot workspace. Ensure the `--workdir` flag points to your workspace directory.

```bash
npx --yes clawhub@latest list --workdir ~/.nanobot/workspace
```

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

### Install Geneclaw with Dashboard Support

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Install Geneclaw with additional dependencies required for the dashboard functionality.

```bash
pip install -e ".[dev,dashboard]"
```

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

### Geneclaw Dashboard Options Example

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Launches the Geneclaw dashboard on a specific port and directs it to load event and benchmark data from specified files.

```bash
nanobot geneclaw dashboard \
  --port 8501 \
  --events /path/to/events.jsonl \
  --benchmarks /path/to/benchmarks.jsonl
```

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

### Run Nanobot in a Container

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Use Docker to run Nanobot in an isolated environment. This example shows how to pull a Python image, install Nanobot, and run it interactively.

```bash
# Run in a container or VM
docker run --rm -it python:3.11
pip install nanobot-ai
```

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

### Verify Geneclaw Setup and Generate Proposal

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

After configuring your LLM provider, use these commands to verify the setup, generate run events, and create a real proposal. The `geneclaw doctor` command is crucial for checking all configurations.

```bash
nanobot geneclaw doctor          # verify everything is green
nanobot agent -m "Hello"         # generate run events
nanobot geneclaw evolve --dry-run  # generate a real proposal
```

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

### Geneclaw Benchmark Options Example

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Sets up benchmark execution with specified event counts, gate iterations, and output format, including saving results.

```bash
nanobot geneclaw benchmark \
  --event-counts 100,500,1000 \
  --gate-iterations 100 \
  --save \
  --format table
```

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

### Update All Installed Skills

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/clawhub/SKILL.md

Updates all skills currently installed in the nanobot workspace. The `--workdir` flag is essential for targeting the correct workspace.

```bash
npx --yes clawhub@latest update --all --workdir ~/.nanobot/workspace
```

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

### Geneclaw Autopilot Options Example

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Configures the autopilot mode with specific parameters for maximum cycles, cooldown, auto-approval, dry-run, and output format.

```bash
nanobot geneclaw autopilot \
  --max-cycles 5 \
  --cooldown 10 \
  --auto-approve low \
  --dry-run \
  --format table
```

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

### Initialize a New Skill

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/skill-creator/SKILL.md

Use the init_skill.py script to create a new skill template. Specify the skill name, output directory, and optionally include resource directories or example files.

```bash
scripts/init_skill.py <skill-name> --path <output-directory> [--resources scripts,references,assets] [--examples]
```

```bash
scripts/init_skill.py my-skill --path skills/public
```

```bash
scripts/init_skill.py my-skill --path skills/public --resources scripts,references
```

```bash
scripts/init_skill.py my-skill --path skills/public --resources scripts --examples
```

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

### Launch Geneclaw Dashboard

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Start the Streamlit dashboard for Geneclaw, accessible at http://localhost:8501.

```bash
nanobot geneclaw dashboard
```

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

### Get Full Weather Forecast (wttr.in)

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/weather/SKILL.md

Retrieve a full weather forecast for a location using wttr.in. Requires no API key.

```bash
curl -s "wttr.in/London?T"
```

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

### Start Isolated tmux Session with Python REPL

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/tmux/SKILL.md

This snippet initializes a new tmux session for a Python REPL, creates a socket for communication, and captures initial pane output. Ensure the NANOBOT_TMUX_SOCKET_DIR environment variable is set or a default temporary directory is used.

```bash
SOCKET_DIR="${NANOBOT_TMUX_SOCKET_DIR:-"${TMPDIR:-/tmp}/nanobot-tmux-sockets}"
mkdir -p "$SOCKET_DIR"
SOCKET="$SOCKET_DIR/nanobot.sock"
SESSION=nanobot-python

tmux -S "$SOCKET" new -d -s "$SESSION" -n shell
tmux -S "$SOCKET" send-keys -t "$SESSION":0.0 -- 'PYTHON_BASIC_REPL=1 python3 -q' Enter
tmux -S "$SOCKET" capture-pane -p -J -t "$SESSION":0.0 -S -200
```

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

### Get Compact Weather Forecast (wttr.in)

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/weather/SKILL.md

Fetch weather data with specific format codes for a compact output. Requires no API key.

```bash
curl -s "wttr.in/London?format=%l:+%c+%t+%h+%w"
# Output: London: ⛅️ +8°C 71% ↙5km/h
```

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

### Get Weather Forecast (Open-Meteo)

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/weather/SKILL.md

Fetch weather forecast data in JSON format from Open-Meteo. This service is free and requires no API key, making it suitable for programmatic use.

```bash
curl -s "https://api.open-meteo.com/v1/forecast?latitude=51.5&longitude=-0.12&current_weather=true"
```

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

### Geneclaw Commit Convention Example

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Follow this commit message format, including the type, scope, subject, and optional metadata like Evo-Event-ID, Risk-Level, and Tests.

```git
feat(geneclaw): add autopilot controller

Evo-Event-ID: abc123
Risk-Level: low
Tests: pytest tests/test_geneclaw_autopilot.py -q
```

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

### Get Current Weather (wttr.in)

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/weather/SKILL.md

Use curl to fetch current weather for a location in a simple format. Requires no API key.

```bash
curl -s "wttr.in/London?format=3"
# Output: London: ⛅️ +8°C
```

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

### Check Working Tree Status

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/upstream-sync.md

Ensure there are no uncommitted changes before starting the sync process. Stash or commit any pending work.

```bash
git status
# Must show: "nothing to commit, working tree clean"
# If not, stash or commit your changes first
```

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

### Check for Vulnerable Dependencies

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Use `pip-audit` to identify vulnerable Python dependencies. Ensure you have the latest secure versions installed.

```bash
# Check for vulnerable dependencies
pip install pip-audit
pip-audit
```

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

### Task Format for HEARTBEAT.md

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/AGENTS.md

Examples of task formats to be used when managing periodic tasks in HEARTBEAT.md. These tasks are checked every 30 minutes.

```markdown
- [ ] Check calendar and remind of upcoming events
- [ ] Scan inbox for urgent emails
- [ ] Check weather forecast for today
```

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

### Phase 1: Bootstrap Allowlist Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/specs/GEP-v0.md

This initial configuration restricts Geneclaw to modifying only its own code and documentation. It's recommended for starting trust-building.

```json
{
  "allowlist_paths": ["geneclaw/", "docs/"],
  "denylist_paths": [".env", "secrets/", ".git/", "config.json", "pyproject.toml"]
}
```

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

### Geneclaw Doctor Check Output

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/quickstart/Geneclaw-Runbook.md

Example output from the `nanobot geneclaw doctor` command, indicating the status of various Geneclaw configuration checks.

```text
Geneclaw Doctor

  ✓ geneclaw.enabled: Geneclaw observability is enabled.
  ✓ dry_run_default: Dry-run is the default (allow_apply_default=false). Safe.
  ✓ restrict_to_workspace: tools.restrictToWorkspace is ON. Recommended.
  ✓ runs_dir_writable: Runs directory does not exist yet. Will be auto-created.
  ✓ allowlist_paths: Allowlist (4 entries): geneclaw/, nanobot/, tests/, docs/
  ✓ denylist_paths: Denylist (4 entries): .env, secrets/, .git/, config.json
  ✓ redact_enabled: Secret redaction is enabled for run logs.

  Summary: 7 ok, 0 warnings, 0 errors

Suggested next steps:
  ...
```

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

### Markdown Structure for PDF Processing Skill

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/skill-creator/SKILL.md

Illustrates a high-level guide pattern for a skill, referencing external files for advanced features like form filling and API reference.

```markdown
# PDF Processing

## Quick start

Extract text with pdfplumber:
[code example]

## Advanced features

- **Form filling**: See [FORMS.md](FORMS.md) for complete guide
- **API reference**: See [REFERENCE.md](REFERENCE.md) for all methods
- **Examples**: See [EXAMPLES.md](EXAMPLES.md) for common patterns
```

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

### Verify Geneclaw Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/first-real-proposal.md

Run the Geneclaw doctor command to check if your configuration is set up correctly. All checks should pass with green indicators for successful setup.

```bash
nanobot geneclaw doctor
```

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

### Download Weather Forecast as PNG (wttr.in)

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/weather/SKILL.md

Save a weather forecast as a PNG image file using wttr.in. Requires no API key.

```bash
curl -s "wttr.in/Berlin.png" -o /tmp/weather.png
```

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

### Pre-Release Checklist Commands

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/release-runbook.md

Execute these commands to ensure the repository is up-to-date, tests pass, and health checks are clear before creating a release.

```bash
git checkout master
git pull origin master

pytest -q

nanobot geneclaw doctor

nanobot geneclaw evolve --dry-run

nanobot geneclaw report

nanobot geneclaw benchmark --format json > benchmark-v0.1.0.json
```

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

### Verify LLM Provider Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

Run the `nanobot status` command to verify that your configured LLM provider is recognized and accessible. A green checkmark indicates a successful connection.

```bash
nanobot status
```

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

### Create GitHub Release using gh CLI

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/release-runbook.md

Use the GitHub CLI to create a new release, linking it to the tag and changelog.

```bash
gh release create v0.1.0 \
  --title "Geneclaw v0.1.0 — GEP v0 Initial Release" \
  --notes-file CHANGELOG.md \
  --latest
```

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

### Add Recurring Cron Reminder

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Adds a recurring reminder using the nanobot cron command. Examples show daily and hourly scheduling.

```bash
# Every day at 9am
nanobot cron add --name "morning" --message "Good morning! ☀️" --cron "0 9 * * *"

# Every 2 hours
nanobot cron add --name "water" --message "Drink water! 💧" --every 7200
```

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

### Run Nanobot as a Dedicated User

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Create a dedicated user account for Nanobot and run the gateway process as that user to limit privileges. Ensure the user has appropriate shell access.

```bash
sudo useradd -m -s /bin/bash nanobot
sudo -u nanobot nanobot gateway
```

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

### Run Post-Sync Verification Suite

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/upstream-sync.md

Execute a series of checks after merging upstream changes to ensure system stability and correctness. This includes running the test suite, performing a doctor check, comparing benchmarks, and verifying the evolve pipeline and report generation.

```bash
# 1. Test suite
pytest -q
# Expected: all tests pass (including geneclaw tests)
```

```bash
# 2. Doctor check
nanobot geneclaw doctor
# Expected: 0 errors
```

```bash
# 3. Benchmark comparison
nanobot geneclaw benchmark --format json > benchmark-post-sync.json
# Compare with previous baseline — significant regressions need investigation
```

```bash
# 4. Evolve pipeline check
nanobot geneclaw evolve --dry-run
# Expected: completes without error
```

```bash
# 5. Report check
nanobot geneclaw report
# Expected: displays without error
```

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

### Directory Structure for Variant-Specific Skill Organization

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/skill-creator/SKILL.md

Demonstrates a file organization pattern for skills supporting multiple frameworks or variants, separating provider-specific details into distinct files.

```directory
cloud-deploy/
├── SKILL.md (workflow + provider selection)
└── references/
    ├── aws.md (AWS deployment patterns)
    ├── gcp.md (GCP deployment patterns)
    └── azure.md (Azure deployment patterns)
```

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

### Create and Push Git Tag

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/release-runbook.md

Create an annotated Git tag for the release and push it to the remote repository.

```bash
# Create annotated tag
git tag -a v0.1.0 -m "Geneclaw v0.1.0 — GEP v0 initial release

Features:
- Observability: JSONL run event recorder with secret redaction
- Evolver: heuristic + LLM-assisted evolution proposals
- Gatekeeper: 5-layer safety validation
- Apply: git-branched diff application with auto-rollback
- Autopilot: multi-cycle evolution loop controller
- Benchmarks: pipeline performance measurement
- Doctor: read-only health checks
- Report: aggregated pipeline statistics (table + JSON)
- CI/CD: GitHub Actions workflow + PR template

54 tests passing. All dry-run by default."

# Push the tag
git push origin v0.1.0
```

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

### Write File Tool

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Writes content to a file, creating parent directories if they do not exist.

```python
write_file(path: str, content: str) -> str
```

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

### Verify Git Remotes

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/upstream-sync.md

Check if both 'origin' (your fork) and 'upstream' (original nanobot) remotes are configured correctly.

```bash
git remote -v
```

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

### Reset to Pre-Merge State

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/upstream-sync.md

If a merge has started but not yet committed, this command can be used to reset the working directory and index to the state before the merge began. Use with caution as it discards uncommitted changes.

```bash
git reset --hard HEAD
```

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

### Orchestrate Multiple Coding Agents with tmux

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/tmux/SKILL.md

This snippet demonstrates how to set up and manage multiple tmux sessions for parallel execution of coding agents. It includes creating sessions, sending commands to different agents, and polling for completion.

```bash
SOCKET="${TMPDIR:-/tmp}/codex-army.sock"

# Create multiple sessions
for i in 1 2 3 4 5; do
  tmux -S "$SOCKET" new-session -d -s "agent-$i"
done

# Launch agents in different workdirs
tmux -S "$SOCKET" send-keys -t agent-1 "cd /tmp/project1 && codex --yolo 'Fix bug X'" Enter
tmux -S "$SOCKET" send-keys -t agent-2 "cd /tmp/project2 && codex --yolo 'Fix bug Y'" Enter

# Poll for completion (check if prompt returned)
for sess in agent-1 agent-2; do
  if tmux -S "$SOCKET" capture-pane -p -t "$sess" -S -3 | grep -q "❯"; then
    echo "$sess: DONE"
  else
    echo "$sess: Running..."
  fi
done

# Get full output from completed session
tmux -S "$SOCKET" capture-pane -p -t agent-1 -S -500
```

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

### Monitor and Capture tmux Pane Output

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/tmux/SKILL.md

Commands to monitor an active tmux session and capture its pane output. Use these after starting a session to observe its state and retrieve recent history.

```bash
To monitor:
  tmux -S "$SOCKET" attach -t "$SESSION"
  tmux -S "$SOCKET" capture-pane -p -J -t "$SESSION":0.0 -S -200
```

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

### Get PR Details with `gh api`

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/github/SKILL.md

Accesses specific fields of a pull request using the `gh api` command with `--jq` for filtering. This is useful for retrieving data not available through standard subcommands.

```bash
gh api repos/owner/repo/pulls/55 --jq '.title, .state, .user.login'
```

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

### Geneclaw CLI Reference - Apply Proposal File Command

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Applies a previously saved evolution proposal from a JSON file.

```bash
nanobot geneclaw apply <file.json>
```

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

### Geneclaw CLI Reference - Benchmark Command

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Runs pipeline performance benchmarks.

```bash
nanobot geneclaw benchmark
```

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

### Example of Geneclaw Redaction

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

Illustrates how Geneclaw's redaction layer masks sensitive information like API keys in logs before they are written. This protects keys even if they are accidentally exposed in conversations or tool outputs.

```text
Input:  api_key="sk-or-v1-abc123def456"
Output: api_key="[REDACTED]"
```

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

### Geneclaw CLI Reference - Evolve Apply Command

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Generates and applies an evolution proposal. Requires specific configuration settings to be enabled.

```bash
nanobot geneclaw evolve --apply
```

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

### Configure Channel Access Control

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Define `allowFrom` lists in the configuration to restrict access to specific user IDs or phone numbers for production environments. An empty list allows all access.

```json
{
  "channels": {
    "telegram": {
      "enabled": true,
      "token": "YOUR_BOT_TOKEN",
      "allowFrom": ["123456789", "987654321"]
    },
    "whatsapp": {
      "enabled": true,
      "allowFrom": ["+1234567890"]
    }
  }
}
```

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

### Phase 3: Full Development Scope Allowlist Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/specs/GEP-v0.md

This configuration allows modifications to the core nanobot framework, tests, and documentation. It still denies access to CI/CD workflows and critical CLI entry points.

```json
{
  "allowlist_paths": ["geneclaw/", "nanobot/", "tests/", "docs/"],
  "denylist_paths": [".env", "secrets/", ".git/", "config.json", "pyproject.toml",
                     ".github/", "nanobot/cli/"]
}
```

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

### View GitHub Release

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/release-runbook.md

Display information about a specific GitHub release using the GitHub CLI.

```bash
gh release view v0.1.0
```

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

### Chat with Nanobot Agent

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Use this command to interact with the Nanobot agent and inquire about available tools.

```bash
nanobot agent -m "Hello, what tools do you have?"
```

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

### View Geneclaw Proposals

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/quickstart/Geneclaw-Runbook.md

Commands to list and read Geneclaw evolution proposals. Use 'ls' or 'dir' to list files and 'cat' or 'type' to read a specific proposal's JSON content.

```bash
# List proposals
ls ~/.nanobot/workspace/geneclaw/proposals/
# Windows: dir %USERPROFILE%\.nanobot\workspace\geneclaw\proposals\
```

```bash
# Read a proposal
cat ~/.nanobot/workspace/geneclaw/proposals/proposal_YYYYMMDD_HHMMSS.json
# Windows: type %USERPROFILE%\.nanobot\workspace\geneclaw\proposals\proposal_...json
```

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

### Search for Skills with ClawHub

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/clawhub/SKILL.md

Use this command to find skills in the ClawHub registry. Specify a search query and an optional limit for the number of results.

```bash
npx --yes clawhub@latest search "web scraping" --limit 5
```

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

### Configure LLM Provider API Key in config.json

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

Set the API key for your chosen LLM provider within the `~/.nanobot/config.json` file. This file is stored locally and excluded from version control to maintain security.

```json
{
  "providers": {
    "openrouter": {
      "apiKey": "sk-or-v1-your-key-here"
    }
  },
  "agents": {
    "defaults": {
      "model": "openrouter/anthropic/claude-opus-4-5"
    }
  }
}
```

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

### Export API Key as Environment Variable (Windows CMD)

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

Set your LLM provider's API key as an environment variable in the Windows command prompt (cmd). This is a temporary setting for the current session. For persistence, add it to your batch file or system environment variables.

```cmd
set OPENROUTER_API_KEY=sk-or-v1-your-key-here
```

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

### Geneclaw Data Layout Overview

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Runtime data for Geneclaw is organized under `~/.nanobot/workspace/geneclaw/`. This includes run logs, evolution events, generated proposals, and benchmark results.

```bash
~/.nanobot/workspace/geneclaw/
├── runs/                      # Run event logs (per session, per day)
│   └── <session_key>/
│       └── YYYYMMDD.jsonl
├── events/                    # Evolution lifecycle events
│   └── events.jsonl
├── proposals/                 # Generated proposals
│   └── proposal_YYYYMMDD_HHMMSS.json
└── benchmarks/                # Performance benchmark results
    └── benchmarks.jsonl
```

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

### Geneclaw CLI Evolve Command (Apply)

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/specs/GEP-v0.md

Generates a proposal and applies it directly. Requires the --apply flag.

```bash
# Generate and apply
nanobot geneclaw evolve --apply
```

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

### Directory Structure for Domain-Specific Skill Organization

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/skill-creator/SKILL.md

Shows a file organization pattern for skills with multiple domains, where content is separated into domain-specific files for efficient loading.

```directory
bigquery-skill/
├── SKILL.md (overview and navigation)
└── reference/
    ├── finance.md (revenue, billing metrics)
    ├── sales.md (opportunities, pipeline)
    ├── product.md (API usage, features)
    └── marketing.md (campaigns, attribution)
```

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

### Geneclaw CLI Apply Command

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/specs/GEP-v0.md

Applies a previously saved proposal file. Requires the --apply flag.

```bash
# Apply a saved proposal file
nanobot geneclaw apply proposal.json --apply
```

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

### Set Secure File Permissions

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Apply strict file permissions to sensitive directories and configuration files. `chmod 700` restricts access to the owner for directories, while `chmod 600` does the same for files.

```bash
chmod 700 ~/.nanobot
chmod 600 ~/.nanobot/config.json
chmod 700 ~/.nanobot/whatsapp-auth
```

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

### Background Tasks

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Tool for spawning background tasks.

```APIDOC
## spawn

### Description
Spawn a subagent to handle a task in the background.

### Signature
```python
spawn(task: str, label: str = None) -> str
```

### Notes
Use for complex or time-consuming tasks that can run independently. The subagent will complete the task and report back when done.
```

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

### Search for Access Denied Logs

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Use `grep` to search the Nanobot log file for 'Access denied' messages, which can indicate unauthorized access attempts. This is a key step in incident response.

```bash
grep "Access denied" ~/.nanobot/logs/nanobot.log
```

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

### Enable Geneclaw in Nanobot Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Add or merge this JSON configuration into your `~/.nanobot/config.json` file to enable Geneclaw and set its parameters.

```json
{
  "geneclaw": {
    "enabled": true,
    "redactEnabled": true,
    "allowApplyDefault": false,
    "allowlistPaths": ["geneclaw/", "docs/"],
    "denylistPaths": [".env", "secrets/", ".git/", "config.json"],
    "maxPatchLines": 500
  }
}
```

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

### List Directory Tool

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Lists the contents of a specified directory.

```python
list_dir(path: str) -> str
```

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

### Launch Dashboard with Custom Data Files

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/dashboard-runbook.md

Launches the Geneclaw dashboard and specifies custom paths for the events and benchmarks data files. This is useful for using pre-generated or specific data sets.

```bash
nanobot geneclaw dashboard \
  --events /path/to/events.jsonl \
  --benchmarks /path/to/benchmarks.jsonl
```

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

### Web Fetch Tool

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Fetches and extracts the main content from a URL, supporting markdown or plain text extraction. Output is truncated by default.

```python
web_fetch(url: str, extractMode: str = "markdown", maxChars: int = 50000) -> str
```

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

### Geneclaw CLI Reference - Benchmark Save Command

Source: https://github.com/clawland-ai/geneclaw/blob/master/README.md

Runs benchmarks and persists the results to a JSONL file.

```bash
nanobot geneclaw benchmark --save
```

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

### Monitor Nanobot Logs

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Use `tail -f` to continuously monitor the Nanobot log file for real-time activity and potential issues. Ensure log files are stored securely.

```bash
# Configure log monitoring
tail -f ~/.nanobot/logs/nanobot.log
```

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

### Manual Geneclaw Proposal Application

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/first-real-proposal.md

This sequence of commands allows for manual application of a geneclaw proposal, providing more control over branching, diff application, testing, and committing.

```bash
# Create a branch
git checkout -b feat/first-real-proposal

# Apply the diff manually
git apply path/to/extracted.patch

# Run tests
pytest -q

# If all pass, commit
git add -A
git commit -m "feat(geneclaw): <proposal title>

Evo-Event-ID: <proposal-id>
Risk-Level: low
Tests: pytest -q"

# Push and open PR
git push -u origin feat/first-real-proposal
gh pr create --title "<proposal title>" --body "Generated by geneclaw evolve"
```

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

### Package Skill Script

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/skill-creator/SKILL.md

Use this script to package a developed skill into a distributable .skill file. It automatically validates the skill before packaging.

```bash
scripts/package_skill.py <path/to/skill-folder>
```

```bash
scripts/package_skill.py <path/to/skill-folder> ./dist
```

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

### Export API Key as Environment Variable (Linux/macOS)

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/llm-provider-setup.md

Set your LLM provider's API key as an environment variable on Linux or macOS. This is a temporary setting for the current session. For persistence, add it to your shell profile.

```bash
export OPENROUTER_API_KEY="sk-or-v1-your-key-here"
```

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

### View Geneclaw Evolution Report

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/quickstart/Geneclaw-Runbook.md

Generate an evolution report for Geneclaw runs from the last 48 hours using the `nanobot geneclaw report` command.

```bash
nanobot geneclaw report --since 48
```

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

### Add a Scheduled Reminder

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/AGENTS.md

Use this command to schedule a reminder for a specific time. Ensure USER_ID and CHANNEL are obtained from the current session.

```shell
nanobot cron add --name "reminder" --message "Your message" --at "YYYY-MM-DDTHH:MM:SS" --deliver --to "USER_ID" --channel "CHANNEL"
```

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

### Geneclaw Evolve Dry-Run Output

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/first-live-run-2026-02-18.md

Output from `nanobot geneclaw evolve --dry-run`, showing the heuristic-only evolution proposal generated in the absence of an LLM provider. It details the analysis of recent events and the resulting proposal's objective and risk.

```text
No LLM provider configured — running heuristic-only mode.
Analysing recent events...

Evolution Proposal: heuristic-only
  ID:         b931ef32-f329-4fab-b5a2-85fece2d8e66
  Objective:  Heuristic diagnosis (no LLM): Top failing tools: web_search(2),
              exec_command(1); Exception clusters: "TimeoutError: web search
              timed out after 30s"(1), "ConnectionError: DNS lookup failed for
              search.api"(1), "CalledProcessError: command exited with code 1"(1)
  Risk:       low
  Files:      (none)
  Evidence:   Top failing tools: web_search(2), exec_command(1); ...

Proposal written to ~/.nanobot/workspace/geneclaw/proposals/proposal_20260218_013923.json

Dry-run mode. Use --apply to apply the proposal.
```

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

### Run Geneclaw Autopilot with Auto-Apply

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/github-governance.md

Enables automatic application of changes after successful dry-runs and meeting specific prerequisites. The 'low' risk setting is recommended for auto-approval.

```bash
nanobot geneclaw autopilot --apply --auto-approve low --max-cycles 1
```

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

### Check for Security Updates

Source: https://github.com/clawland-ai/geneclaw/blob/master/SECURITY.md

Regularly check for and apply updates to the `nanobot-ai` package to ensure you have the latest security patches. This is crucial for maintaining a secure deployment.

```bash
# Check for updates weekly
pip install --upgrade nanobot-ai
```

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

### Export OpenRouter API Key

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/first-real-proposal.md

Set your OpenRouter API key as an environment variable. This is necessary for the LLM provider to authenticate. Ensure the key is also configured in your Nanobot config file.

```bash
# Linux/macOS
export OPENROUTER_API_KEY="sk-or-v1-your-key-here"
```

```powershell
# Windows PowerShell
$env:OPENROUTER_API_KEY = "sk-or-v1-your-key-here"
```

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

### Scheduled Reminders (Cron)

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Instructions for managing scheduled reminders using the `nanobot cron` command.

```APIDOC
## Set a recurring reminder

### Example
```bash
# Every day at 9am
nanobot cron add --name "morning" --message "Good morning! ☀️" --cron "0 9 * * *"

# Every 2 hours
nanobot cron add --name "water" --message "Drink water! 💧" --every 7200
```
```

```APIDOC
## Set a one-time reminder

### Example
```bash
# At a specific time (ISO format)
nanobot cron add --name "meeting" --message "Meeting starts now!" --at "2025-01-31T15:00:00"
```
```

```APIDOC
## Manage reminders

### Commands
- `nanobot cron list` - List all jobs
- `nanobot cron remove <job_id>` - Remove a job
```

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

### Configure Geneclaw in nanobot

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/quickstart/Geneclaw-Runbook.md

Edit the nanobot configuration file to enable Geneclaw and set various operational parameters like redaction, path allowlists, and denylists.

```json
{
  "geneclaw": {
    "enabled": true,
    "redactEnabled": true,
    "allowApplyDefault": false,
    "logMaxChars": 500,
    "maxPatchLines": 500,
    "allowlistPaths": ["geneclaw/", "nanobot/", "tests/", "docs/"],
    "denylistPaths": [".env", "secrets/", ".git/", "config.json"]
  },
  "tools": {
    "restrictToWorkspace": true
  }
}
```

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

### Push Changes to Origin

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/upstream-sync.md

Push the synchronized 'master' branch to your 'origin' remote repository.

```bash
git push origin master
```

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

### Launch Dashboard on Custom Port

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/ops/dashboard-runbook.md

Launches the Geneclaw dashboard on a specified port, useful if the default port 8501 is already in use.

```bash
nanobot geneclaw dashboard --port 9090
```

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

### Phase 2: Expanded Allowlist Configuration

Source: https://github.com/clawland-ai/geneclaw/blob/master/docs/specs/GEP-v0.md

After successful evolution cycles, expand the allowlist to include tests. This allows Geneclaw to propose test improvements and new test cases.

```json
{
  "allowlist_paths": ["geneclaw/", "docs/", "tests/"],
  "denylist_paths": [".env", "secrets/", ".git/", "config.json", "pyproject.toml"]
}
```

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

### File Operations

Source: https://github.com/clawland-ai/geneclaw/blob/master/workspace/TOOLS.md

Tools for interacting with the file system.

```APIDOC
## read_file

### Description
Read the contents of a file.

### Signature
```python
read_file(path: str) -> str
```
```

```APIDOC
## write_file

### Description
Write content to a file (creates parent directories if needed).

### Signature
```python
write_file(path: str, content: str) -> str
```
```

```APIDOC
## edit_file

### Description
Edit a file by replacing specific text.

### Signature
```python
edit_file(path: str, old_text: str, new_text: str) -> str
```
```

```APIDOC
## list_dir

### Description
List contents of a directory.

### Signature
```python
list_dir(path: str) -> str
```
```

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

### Search Past Events with Grep

Source: https://github.com/clawland-ai/geneclaw/blob/master/nanobot/skills/memory/SKILL.md

Use this command to search for keywords within the event history log. Case-insensitive matching is enabled.

```bash
grep -i "keyword" memory/HISTORY.md
```