For AI agents · MCP server

Your forms, agent-readable.

Connect Claude, ChatGPT, Cursor, Windsurf, or Zed to your Formspring workspace in 30 seconds. Every form, every submission, every webhook — exposed as tools an agent can call directly. Tokens are scoped, ability-gated, and team-isolated.

Read the MCP guide → Quick install

Tools

Mutating actions across forms, submissions, webhooks, billing, insights, tokens.

Resources

Read-only `formspring://` URIs the agent can subscribe to.

Prompts

Turn-key playbooks (triage, audit, draft, summarize).

Abilities

Mint a token with exactly the abilities the agent needs — no more.

Pricing

Pro+

Bundled with API access. No per-call charges.

Why MCP

The difference between a chatbot that talks about forms and one that runs them.

MCP lets agents read your inbox, mark spam, draft autoresponders, audit webhooks, and ship configuration changes — all from inside Claude or Cursor. No glue code. No sync jobs. No screen scraping.

Triage from anywhere

Ask your agent to review spam, mark false positives, and patch your form's custom rules — all in one chat session.

Draft like a teammate

Hand the agent a form's intent and let it draft autoresponder copy, propose new fields, or test-send before publishing.

Scoped by default

Every token carries a list of abilities. Read-only? Read + spam triage? You decide. The agent only sees what your token allows.

Same shape as the API

Tools mirror the REST API one-for-one. If you can do it in the dashboard, you can do it from an agent.

Webhooks as a conversation

Audit recent deliveries, replay failures, and update destinations without leaving your editor.

Honest documentation

Every tool, every ability, every error code is in the docs. No black boxes.

Install

30 seconds to connect.

Mint a token in your account, drop it into your client config, and your agent has tools.

Pick your client

claude mcp add formspring \
  --transport http \
  --url https://formspring.io/mcp-server \
  --header "Authorization: Bearer your-api-token"

Mint a token at your account API page. Pick the abilities you want the agent to have.

Catalogue

33 tools. 9 resources. 8 prompts.

Tools mutate. Resources are read-only views with stable URIs. Prompts are turn-key playbooks the agent runs end-to-end. Every entry lists the ability it requires.

Forms

list_forms
get_form
create_form
update_form
publish_form
unpublish_form
archive_form
restore_form
send_test_autoresponder
send_test_notification

Submissions

list_submissions
get_submission
update_submission
delete_submission
bulk_submissions
export_submissions
get_submission_file
mark_as_spam
get_form_stats

Webhooks

list_webhooks
get_webhook
create_webhook
update_webhook
delete_webhook
list_webhook_deliveries
replay_webhook_delivery

Billing (read-only)

get_team
get_plan
get_usage
get_subscription_state

AI Insights

get_ai_insights
regenerate_ai_insights

Tokens (self-management)

list_tokens
create_token
revoke_token

Resources · 9 URIs

Read-only views with stable URIs.

formspring://team/forms

List of active forms

formspring://forms/{public_id}

Single form

formspring://forms/{public_id}/submissions/recent

Last 25 inbox submissions

formspring://forms/{public_id}/stats

30-day stats

formspring://forms/{public_id}/insights

Latest AI digest

formspring://team/usage

Current period usage

formspring://team/plan

Plan, features, limits

formspring://webhooks/{webhook_id}

Single webhook

formspring://webhooks/{webhook_id}/deliveries/recent

Last 50 deliveries

Prompts · 8 playbooks

Turn-key playbooks the agent runs end-to-end.

triage-spam

Review spam, propose corrective rules

summarize-week

Last-7-day exec summary for a form

draft-autoresponder

Draft + test + confirm before saving

audit-webhooks

Bucket failures, propose patches

categorize-submissions

Apply taxonomy via update_submission

draft-form

Sketch a new form schema, then create_form

weekly-digest-brief

Read team digest, recommend actions

integration-health

Per-team webhook health card

Paste-ready prompts

One sentence to your agent. One tool call to your forms.

Copy any of these into Claude, Cursor, Windsurf, or Zed. They work as-is — replace the form name with one of yours.

Triage spam

submissions:read · submissions:write

“Triage today’s spam folder for the Contact form. Walk me through each submission, decide keep / mark not-spam / delete, and apply via bulk_submissions when I confirm.”

Audit webhooks

webhooks:read · webhooks:write

“Audit every webhook on every form. For each, check the last 100 deliveries. Tell me which are unhealthy and why. Sort the report worst-first.”

Draft autoresponder

forms:read · forms:test-email · forms:write

“Draft an autoresponder for the Contact form. Friendly tone. Mention our 24-hour response window. Send a test to my account email — don’t save until I confirm.”

Weekly digest

submissions:read · insights:read

“Build me a weekly digest for the Contact form. One headline number, spam ratio movement vs. last week, top 3 themes, one outlier. Markdown, ready to paste into email.”

Categorize backlog

submissions:read · submissions:write

“Re-categorize every uncategorized submission on the Contact form from the last 30 days. Use the form’s existing categories. Show me 5 sample decisions before applying the rest.”

Health check

forms:read · webhooks:read

“Run an integration health check across the team. Flag every form whose autoresponder is enabled but hasn’t fired, and every webhook with <90% success rate.”

Tokens are the fence.

Tokens are scoped, ability-gated, and team-isolated. Agents only see what your token allows. Revoke any time from the API page.

Scoping

Every tool checks the token has the right ability. No ability, no call.

Team isolation

Tokens are minted in a single team. Cross-team reads return "not found".

Revocable

Pull a token in one click. Agents lose access immediately on the next call.

Ship your form in two minutes.

No credit card. 50 free submissions a month, every month.

Start free Read the docs