Word Is Bond
Start here

What it does, what a run looks like, how to run one

Three steps, in order. Read the first, open the second, follow the third. Nothing on this page needs an account until step three.

Open a scored run Read the docs

1. What Word Is Bond does

Word Is Bond places scripted test calls to a voice agent you control, then scores what came back. A synthetic caller with a stated goal and persona talks to your agent, and the run produces four separate signals rather than one opinion: audio quality and latency, whether the agent stayed on script, how it handled interruption and tone, and whether the caller's goal was actually completed.

The point is the comparison. Every run is scored the same way, so a prompt edit, a model swap, or a changed integration shows up as a movement in a number instead of a report from an annoyed customer three weeks later.

2. What a scored run looks like

The scorecard is the artifact. It carries the pass or fail, the four layer scores, the measured metrics, a transcript excerpt, and a regression note explaining what moved since the last passing baseline.

Open the public sample report — a full run against a hosted demo agent. No account, no customer data, no retained call audio.

3. How to run one yourself

The free tier needs no credit card and includes 60 test minutes a month over direct or SIP transport.

  1. Create an account and open the getting-started checklist.
  2. Score a sample transcript — this returns a real scorecard before you connect anything.
  3. Add a target: the SIP URI, WebRTC endpoint, or number of an agent you control.
  4. Pick a persona and a goal, run the call, and open the scorecard.

Once you have an API key, the same run starts from the command line, and the exit status is what you wire into a build: 

# Run a suite and fail the build if it regressed
runId=$(curl -s -X POST https://api.wordis-bond.com/api/tests/run \
  -H "Authorization: Bearer $WORDIS_BOND_KEY" \
  -H "Content-Type: application/json" \
  -d '{"suiteId":"<your-suite-id>"}' | jq -r .data.id)

curl -s https://api.wordis-bond.com/api/tests/$runId \
  -H "Authorization: Bearer $WORDIS_BOND_KEY" | jq -e '.data.passed == true'

Full endpoint reference, scheduled runs, and the ownership boundary are in the docs. If you run your agent on a platform we already speak to, start from its integration page.

Where new walkthroughs land

Recorded runs and walkthroughs are published to the Word Is Bond channel on Capricast.