crawl4ai-install

Install Crawl4AI runtime

Before asking for confirmation, clearly tell the user:

• installation normally takes about 5–15 minutes, depending on internet speed, the Python package cache, and the computer;
• downloads may exceed 500 MB because Crawl4AI dependencies and Chromium binaries are included;
• approximately 2 GB of free disk space is recommended;
• Python 3.10+, venv, network access, and supported OS libraries are required;
• the terminal may appear busy for several minutes during large package or browser downloads.

Ask for confirmation before downloading anything. Then select one command:

# Persistent plugin data, shared across projects (recommended)
"${CLAUDE_PLUGIN_ROOT}/bin/crawl4ai" install --scope custom --directory "${CLAUDE_PLUGIN_DATA}" --project-root "${CLAUDE_PROJECT_DIR}"

# Current project only
"${CLAUDE_PLUGIN_ROOT}/bin/crawl4ai" install --scope project --project-root "${CLAUDE_PROJECT_DIR}"

# Explicit directory
"${CLAUDE_PLUGIN_ROOT}/bin/crawl4ai" install --scope custom --directory "/absolute/path" --project-root "${CLAUDE_PROJECT_DIR}"

Run the selected command directly so its [step/total] progress messages remain visible; do not hide all output behind tail or claim that a quiet step is stalled. The installer prints elapsed time and announces each environment, dependency, verification, and browser stage.

The installer creates separate .venv and .trafilatura-venv environments to avoid dependency conflicts. It runs crawl4ai-setup when available. Do not claim browser readiness unless the command confirms it. Exit code 2 means Python packages installed but browser setup was not verified. --skip-browser intentionally skips browser installation and must be reported as such.