Draft a blog post from what customers ask on calls
Reads recent Gong call transcripts, finds the question customers ask most, and saves a post answering it as a Webflow draft.
Outcome
The question customers asked most on your recent calls, and how many calls raised it.
A blog post answering it, with no customer names, companies or quotes, saved as a draft in your Webflow blog.
How it works
- 1
Read recent calls
Gong
Request the transcripts with
fromDateTimeset tosinceandtoDateTimeset to now, following thecursoruntil none is returned. Keep each call'scallIdand its sentences. - 2
Find the question
Your agentAcross the transcripts, find the product questions customers ask, group the ones that mean the same thing, and pick the one raised on the most calls. Keep the question, how many calls raised it, and the answers your team gave.
- 3
Write the post
Your agentDraft a blog post that answers the question in full, building on the answers your team gave. Leave out every customer's name, company and words. Show the draft to the user, and keep the approved title, slug and body.
- 4
Save the draft
WebflowAfter the user approves, read the fields of
blog_collection, then create one item in it withisDraftset totrue, filling its name, slug and body from the approved post. Keep the item's ID.
You'll be asked for
How far back to read calls
e.g. 30 days ago
The Webflow CMS collection your blog posts live in, by name
e.g. Blog Posts
The file your agent runs
customer-calls-to-blog-post.md
Draft a blog post from what customers ask on calls
Reads recent Gong call transcripts, finds the question customers ask most, and saves a post answering it as a Webflow draft.
Set up the tools below, then run the steps in order for the user, carrying each step's results into the next. The run is done when the user has the outcome below.
Outcome
- The question customers asked most on your recent calls, and how many calls raised it.
- A blog post answering it, with no customer names, companies or quotes, saved as a draft in your Webflow blog.
Inputs
Ask the user for these before you start.
since: how far back to read calls, e.g. 30 days agoblog_collection: the Webflow CMS collection your blog posts live in, by name, e.g. Blog Posts
Set up
Get call transcripts (Gong, tool:gong/get-call-transcripts)
Use the API.
- Base URL: https://{company}.api.gong.io
- Endpoint:
POST /v2/calls/transcript - Auth: send the header
Authorization: Basic $GONG_API_KEY - Get a key: https://app.gong.io/company/api
Note: {company} is your API host prefix from Gong's API settings, as in company-17.api.gong.io. GONG_API_KEY holds the base64 of <access key>:<secret>; 3 calls a second, 10,000 a day.
Note: When more records remain, repeat the request with the returned cursor. OAuth apps need the api:calls:read:transcript scope.
Create CMS items (Webflow, tool:webflow/create-cms-items)
Use the first option your agent supports.
Note: Read the collection's fields first (get_collection_details on MCP). Setting isDraft to false only queues items for the next site publish: publish them to go live now. The 100-item limit counts every locale variant.
MCP (official, remote)
Add this server to your agent's MCP settings, then sign in when asked.
{ "mcpServers": { "webflow": { "url": "https://mcp.webflow.com/mcp" } } }
Call the MCP tool data_cms_tool with POST /collections/{collection_id}/items/insert.
CLI (official)
Install the command and sign in with it, then confirm it runs.
npm install -g @webflow/webflow-cli
webflow --version
Run webflow cms items create.
Before step 1, confirm access to each service with its cheapest read-only call, like a list or a search. Never send or change anything to test access.
Steps
- Read recent calls with Get call transcripts (Gong). Request the transcripts with
fromDateTimeset tosinceandtoDateTimeset to now, following thecursoruntil none is returned. Keep each call'scallIdand its sentences. - Find the question yourself. Across the transcripts, find the product questions customers ask, group the ones that mean the same thing, and pick the one raised on the most calls. Keep the question, how many calls raised it, and the answers your team gave.
- Write the post yourself. Draft a blog post that answers the question in full, building on the answers your team gave. Leave out every customer's name, company and words. Show the draft to the user, and keep the approved title, slug and body.
- Save the draft with Create CMS items (Webflow). After the user approves, read the fields of
blog_collection, then create one item in it withisDraftset totrue, filling its name, slug and body from the approved post. Keep the item's ID.
Notes
The post stays a draft: publish it from Webflow once someone has reviewed it. Transcripts can hold private details, so the post answers the question in your team's words, never a customer's.
Rules
- Use only the services set up above. The read-only calls they need, like listing ids or polling for results, are fine.
- Ask the user before anything that sends messages, costs money, or changes data, and say how many records it touches. One approval covers a batch the user has seen.
- Never print API keys.
