Configuration
config.toml, environment variables, storage, and JSON output.
config.toml
Create config.toml in your platform's config directory:
| Platform | Path |
|---|---|
| macOS | ~/Library/Application Support/omniscout/config.toml |
| Linux | ~/.config/omniscout/config.toml |
search_limit = 10
research_results = 8
request_throttle_seconds = 1.0
browser = "chrome" # chrome | edge | brave | vivaldi | opera | arc | chromium | custom
# browser_executable = "/path/to/browser" # only needed for 'custom'Prefer the CLI over hand-editing — omniscout settings set writes the same file:
omniscout settings set browser brave
omniscout settings showCheck install state with omniscout install, omniscout settings browsers, and omniscout settings engines.
Environment variables
| Variable | Purpose |
|---|---|
OMNISCOUT_JSON=1 | Force JSON output on every command |
OMNISCOUT_DATA_DIR | Override the data directory |
OMNISCOUT_BROWSER | Browser id (overrides config) |
TWOCAPTCHA_API_KEY / CAPTCHA_API_KEY | CAPTCHA solver key |
OMNISCOUT_API_KEY | Cloud API key (Cloud SDK/CLI use) |
OMNISCOUT_BASE_URL | Cloud base URL (default production) |
JSON output
Every command speaks JSON — the agent integration surface:
omniscout --json answer "What is local-first software?"
export OMNISCOUT_JSON=1 # same for all commands in this shellWith JSON on, stdout is a structured payload and logs go to stderr (add -v/--verbose for more). Pipe through jq:
omniscout search "topic" | jq '.hits[] | {title, url}'See Using with AI agents for the full agent loop.
Storage
Saved logins, memory, saved pages, and logs all live in the data dir. Keep them outside version control.