Customers ask support questions on WhatsApp expecting an instant answer, but a generic bot just makes things up. This agent grounds every reply in your actual support docs, so it answers from your policies rather than guessing.
sonnet5 daysWhatsApp Business APIpgvectorClaudeSupabase
Claude
79ROI
76Scale
$5.1k96Saved
ROI for
README.md
Why this subagent
Customers ask support questions on WhatsApp expecting an instant answer, but a generic bot just makes things up. This agent grounds every reply in your actual support docs, so it answers from your policies rather than guessing.
Each response cites the article it came from, and when nothing matches, it routes the customer to a human. Common questions get accurate same-minute answers, and your agents only handle the cases the docs do not cover.
How it runs
Used at step 01 to kick off the pipeline.
Write
Used at step 01 to kick off the pipeline.
WebFetch
Used at step 01 to kick off the pipeline.
WebSearch
Used at step 01 to kick off the pipeline.
Receive the customer's WhatsApp question and retrieve the most relevant passages.
pending
Answer using only those passages and cite the policy or article it came from.
pending
Escalate to an agent when the knowledge base has no good match.
pending
Sample output
json
// Sample output
// (generated when the pipeline finishes)
Answer the customer using only the retrieved support docs; if no passage covers it, reply with an escalation.
Unlock the rest
The full agent definition, install snippet, and starter task are gated for community members.
Members get the full `.md` agent file, the npm / pnpm install one-liners, a starter prompt that we've tuned against real runs, and the open-source repo when this automation ships there. One email, magic link, done.