Is Furrow Forms MCP server safe?

Yes, with the usual care.

A89/100grade A

Passed every safety check we run. Maintained, authenticated, reachable, clean scan.

No critical or high findings in the latest scan.

How Furrow Forms compares

Against every graded AI and agents MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score: 89/100; higher than 83% of 4,460 AI and agents servers
Tools that change things: 31%; typical: 0%. Fewer write tools means less an agent can break
Uptime: 100%; of our recent checks; the typical remote AI and agents server answers 100%

What its tools let an agent do: 16 read · 4 create · 6 change · 3 delete

Public scan report

scanner v0.1.10 · 2026-10-31 · same rubric, same numbers if you re-run it

1 low

Code scan2 source files scanned25/25
Live reliabilityremote reachable in 1105ms20/20
Tool poisoning29 tool descriptions checked13/15
Auth qualityAPI key sent as a header8/15
Maintenancelast push 2 days ago15/15
Maintainer identitynamespace and repository owner differ; GitHub account older than a year; website matches verified namespace8/10

Findings (1)

lowUnusually long tool description (over 2,000 characters)poison.long-description

tool update_account: …Use to rename the workspace, set branding, configure the AI spam filter, or connect the team's own Twilio account for new-lead text alerts. Pass sms { account_sid, auth_token, from } to save Twilio (auth_token is write-only; omit it later to keep the saved one; sms: null disconnects). Numbers to text are notify_phones on the project. US senders must finish A2P 10DLC or toll-free verification in Twilio or carriers drop the texts. To set it up with a human who has never used Twilio, follow furrow://docs/text-alerts. Pass ai.typesafe_api_key to save this team's TypeSafe key (null clears it; the key is write-only). ai.notify_suspected_spam (default true) emails notify addresses with a Potential spam label for every spam catch — rule-based or Jev — except exact duplicates; set it false to silence both. Spam never sends a webhook. Free accounts keep branding.mode and email.footer_mode at furrow; asking for anything else returns plan_limit — tell the human the custom footer is a Yearly feature and offer create_upgrade_link. Yearly may set branding.mode to furrow, off, or agency (name + optional link + optional logo; Furrow keeps the layout, no raw HTML). Plans are capacity, not a feature ladder. Free: one workspace, 100 clean submissions/month pooled on the team, 3 active (non-archived) projects, 30-day submission retention, and email.footer_mode / branding.mode stay furrow (other values return plan_limit). Yearly ($199/year): extra workspaces, 10,000 submissions/month, unlimited projects, 730-day retention, branding.mode furrow | off | agency. Expansion packs are $99/year each and add 5,000 submissions/month on a yearly account. MCP, webhooks, snippets, inheritance, Turnstile, and email templates stay available on free. Spam, honeypot, and rejected posts do not count. Use create_upgrade_link to hand a human a Stripe Checkout (or Customer Portal) URL. Pass product=expansion on a yearly team to add a pack. Operator tokens (all workspaces) must pass team_id from list_teams. Team-scoped tokens may omit team_id and cannot target another workspace. Do not use this to change the plan — call create_upgrade_link.…

Overall 89/100. Components that don't apply are left out of the denominator. Any critical finding is an F.Rubric Appeal a finding JSON

Other servers that do what Furrow Forms does

LLM Sandbox
Securely run LLM-generated code in isolated containers across 7 languages and 3 container backends.
not reviewed
Widely used
A
Agent-Native Forms
Agent-Native form builder - create, edit, and manage forms
not reviewed
Established
A
IWE (iwe.md)
Markdown knowledge base as agent memory. Runs against the notes directory it is started in.
not reviewed
Established
A

Furrow Forms reviews, tools and install