Frequently Asked Questions
What is GTM Tools?
A single MCP service for go-to-market work, organized into five categories: LinkedIn automation, Reddit automation, buying-intent signals, email finding, and centralized auth + billing (Billing). Every tool ships through MCP, REST, and a shared CLI.
How do I get an API key?
Call get_api_key with your email and you’ll get a verification code. Call it again with the code to receive the key. New accounts receive 100 free tokens. See Quickstart.
How is usage metered?
In tokens. $1 = 100 tokens. Each tool’s cost is published via get_token_balance. The cost is debited before the tool runs, so calls rejected up front (400, 401, 402, 404) cost nothing. See How Tokens Work.
Why are some tools 0 tokens?
list_connected_linkedin_accounts, detect_signal, set_signals_order, get_signals_order, and every Billing tool is free. They’re either dispatchers, session helpers, or wallet operations, so they don’t burn provider budget.
detect_signal is free because it’s a dispatcher; the individual detectors it runs cost 5 tokens each.
What's the most expensive tool?
list_linkedin_company_employees_posts (80 tokens). It fans out across an entire company’s employee list. Use sparingly; see Token Efficiency.
Does send_linkedin_invitation actually use my LinkedIn?
Yes. send_linkedin_message and send_linkedin_invitation write to LinkedIn through the session shared by the browser extension. You install it, press Connect, and your own logged-in account performs the writes.
What happens if I run out of tokens?
The next paid tool call returns 402 Insufficient Tokens. No charge applied, no work done. Top up with buy_tokens (one-off) or set set_auto_reload to charge automatically when the balance dips below your threshold.
What's the difference between detect_signal and the individual detectors?
detect_signal runs every detector in the configured order (set via set_signals_order) and returns the union. Use it for fresh accounts. Use individual detectors when you only care about one signal or want to short-circuit on the first hit.
How accurate is get_email?
get_email generates every common pattern and verifies each via SMTP, so an ok result is a genuinely deliverable address. When it can’t confirm one it returns status: "not_found" with a reason (catch_all, no_mx, unverifiable_provider, no_pattern_verified), and the right next step differs per reason. See Why is get_email returning not_found?.
Does GTM Tools send emails?
No. get_email only verifies deliverability: it speaks SMTP up to RCPT TO but never issues DATA. No mail is actually sent.
What boolean syntax do title filters support?
AND, OR, NOT, parentheses, and quoted phrases. Example: "(VP OR Director OR Head) AND \"customer success\" NOT intern".
How do I stop my account from getting rate-limited?
We don’t rate-limit per workspace, so a 429 is either a tool saying try_again_later or an upstream provider throttling us. Honor the Retry-After header, and keep batch concurrency around 5–10 to stay clear of provider-side limits. See Rate Limits.
Is the source available?
Partly. The Agent Skills catalog is public, so you can read and fork every skill. The platform itself (API, tool implementations, extension) is closed source.
Where do I get support?
Email support@gtm-tools.sh for account or billing help, for bugs and missing tools. Skill issues belong on gtm-skills.