The Go-to-Market Agent surfaces high-intent accounts hiding in your self-serve base by analyzing product-usage signals, feature adoption patterns, and engagement frequency. When it identifies an account showing enterprise-level usage on a self-serve plan, it crafts personalized outreach for the most relevant rep and routes the opportunity with full context attached.

Every handoff includes the account's usage timeline, the features driving the signal, comparable closed-won accounts, and a suggested talk track — so reps start the conversation informed rather than cold.

Clone the repo, open it in Warp, and ask the agent to run the setup skill:

Run the setup skill

It installs dependencies, validates the install with the test suites, walks you through .env configuration for the agents you care about, and points you at the per-agent deployment docs (plg_upsell/README.md, bdr_agent/README.md).

Or set it up manually — every variable ships with a placeholder default, so the test suites run cleanly with zero credentials:

python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env

# Validate the install — no credentials required
python3 -m unittest discover -s tests -t .

For live runs, fill in .env: a Google Cloud project with BigQuery (GCP_SERVICE_ACCOUNT_JSON or Application Default Credentials), a HubSpot private app token (HUBSPOT_PRIVATE_APP_TOKEN), and a Slack bot token (PLG_SLACK_BOT_TOKEN) with chat:write scope for the digests.

Connect the agent to your product analytics platform and CRM. Define the usage signals that indicate upgrade intent — feature unlock attempts, seat expansion, heavy API usage, or admin activity spikes.

Configure per-segment routing rules so the right rep type receives each opportunity. Self-serve expansion routes to an account executive, enterprise-usage signals route to enterprise sales, and product-led growth accounts route to a PLG motion sequence.

For production, store secrets in your Oz environment and put the weekly scoring sync on a schedule:

oz schedule create \
  --team \
  --name "plg-hubspot-weekly-sync" \
  --cron "0 11 * * 1" \
  --environment <ENV_ID> \
  --skill "<your-org>/<your-repo>:hubspot-plg-sync" \
  --prompt "Run this week's PLG scoring sync and post the digest."

HubSpot or Salesforce for CRM routing and opportunity creation, Amplitude for usage signal analysis, Slack for real-time rep notifications.

Outreach and Apollo for automatic enrollment into follow-up sequences once the rep accepts the handoff.