Compare

BugHerd vs BugTape: use both

BugHerd collects client feedback pinned to a live page. BugTape catches errors and failed requests users never report and hands them to your agent.

Short answer: BugHerd collects website feedback from people. A client or teammate clicks an element on a live page, leaves a comment, and BugHerd saves a screenshot and browser details as a task on a kanban board. BugTape catches the failures your real users hit and never report. It captures errors, failed requests and requests that hang on its own, groups repeats into one issue, and reopens the issue if the failure comes back in a later release. Both connect to coding agents over MCP. Use BugHerd for client and design review, and BugTape for the bugs nobody tells you about.

No account: npx -y bugtape --help prints the commands and does not sign you in. The demo MCP at https://app.bugtape.ai/demo/mcp needs no sign-in. Your own workspace needs one.

Page owner: Jaco Veldsman, hi@bugtape.ai. Last checked 2026-10-26. I build BugTape. The BugHerd facts come from BugHerd's home page, features page, pricing page, and its help articles on the MCP server and the LogRocket integration, all read on 26 October 2026.

This page is not a ranking. BugTape does not replace BugHerd. BugTape has no pins on page elements, no guest review flow for clients and no kanban board.

The short version

The two tools start from different events. In BugHerd, a person sees something and comments on it. BugHerd's home page says clients "can just point, click & comment on a live website" and do not need an account. In BugTape, nobody has to act: the browser capture on your site files a report when a supported failure happens. Users can still press an optional bug button. If your work is a website build that a client must sign off, you want BugHerd. If your app is live and users leave when something breaks, you want BugTape.

Facts table

	BugHerd	BugTape
Who starts the capture	A person. A client, guest or team member pins a comment on a page element. Visitors to a live site can also report issues through the public feedback widget, on every plan (features, pricing). Install is a one-line JavaScript tag, which BugHerd recommends, or a browser extension when you cannot add code (home page, pricing).	Nobody, for supported failures: uncaught errors, failed requests after a user action, and requests that hang. Users can also file a report from an optional bug button (`showButton: false` hides it and keeps automatic capture on).
Bugs nobody reports	Only if someone notices and leaves feedback, for example through the public feedback widget	Captured automatically when they match a supported failure, with the lead-up
What the evidence looks like	A screenshot with the pin, plus browser, operating system, screen resolution and the element the feedback is pinned to. Short video and audio clips. Feedback on Figma designs, PDFs and images. Features	Browser replay rebuilt from DOM events (not video), the failed request, console and stack. A bug-button report attaches the last ten minutes of page activity. No video, voice or screenshots.
Session replay and console	Its features and pricing pages do not list its own session replay or console logs. If LogRocket is already recording on the page, BugHerd saves a link to that LogRocket session on the feedback. LogRocket integration	Browser issues can include a replay rebuilt from DOM events and the console tail. A report can lack the events needed to rebuild a replay. Servers have no replay or console.
How work is organised	Each piece of feedback becomes a task on a kanban board. Similar task detection (beta) is on Premium and Custom. Pricing	Every report of the same failure becomes one issue, with a count, the releases it hit and affected users when your app sends user IDs
Clients and guests	Unlimited guests on every plan. Clients can review websites and files without logging in (features). A Guest Task Board and white labelling are on Premium and Custom. Pricing	Built for the team that ships the app. Members are 3 on Free and 10 on Pro. There is no guest review flow.
Agent access	BugHerd MCP (beta on its pricing page, all plans). OAuth sign-in. Agents can list and read tasks, create and update tasks, add comments and manage attachments. Named clients include Claude Code, Cursor, ChatGPT, VS Code with GitHub Copilot and Codex CLI. MCP article	MCP at `https://app.bugtape.ai/mcp` with OAuth: `list_recommended_critical`, `get_repro_context`, `ack_bug`. `npx -y bugtape connect cursor` (or `claude`, `codex`). Signed webhooks can wake a headless agent.
After the fix	Update the task on the kanban board	`ack_bug` with `working` and a `prUrl`. A person reviews the PR. If the same failure arrives from a later release, the issue reopens as a regression and your team and agent are told.
Integrations	Trello, Slack, MS Teams, GitHub, FullStory, LogRocket, Zapier, WordPress and more on Standard. Jira, Linear, Asana, ClickUp, Monday.com and Teamwork on Premium and Custom. Pricing	Slack, email, signed webhooks and MCP clients
Price (checked 2026-10-26)	Standard $50 a month (5 members), Studio $80 (10 members), Premium $150 (25 members), Custom on request. Paid yearly: $42, $67 and $125 a month. Unlimited projects and guests. BugHerd pricing. 7-day free trial, no credit card, and a 60-day money-back guarantee (home page). Extra members $8 a month each ($6.60 billed yearly) (pricing).	Free $0, no card: 100 new issues and 150 credits a month. Prepaid credits $10 for 2,000. Pro $29 a month. On Free and Prepaid credits each agent read of a bug costs 2 credits; reads are free on Pro. Pricing

Use both

Your agency builds a site and the client reviews it in BugHerd. The client pins a comment on a button that has the wrong colour on the staging page. Your agent reads the task over BugHerd MCP and makes the change.
The site goes live. A real visitor hits a failed checkout request and leaves. Nobody leaves feedback. BugTape's browser capture files an issue with the request, console and lead-up.
More visitors hit it. BugTape adds them to the same issue, so you can see it is one bug that reaches many people.
Your agent picks it up. It calls get_repro_context, works on the fix, and records the prUrl with ack_bug. A person reviews and merges.
BugTape watches the next release. When you mark the issue fixed, BugTape anchors it to the release. If the same failure comes back from a newer release, the issue reopens as a regression.

BugHerd covers what your client and team point at. BugTape covers what your users hit when nobody is looking.

When BugHerd is the right tool

An agency needs sign-off from a client who will not install anything or make an account.
The feedback is about design, copy or layout. Nothing in the code threw an error or failed a request.
You want comments on Figma files, PDFs and images too.
You want visitors to your live site to report what they see, with a screenshot and browser details.
Your team runs feedback as tasks on a board and links it to Jira, Asana or ClickUp (Premium and above), or to Trello, GitHub or Slack on any plan.

When BugTape is the right tool

Users hit errors, failed requests or hung requests and do not tell you.
The same failure keeps arriving and you want it counted as one issue.
You want the replay, console and failed request with each bug without adding a second replay product.
You want your agent to pick up production bugs over MCP and you want a check after the release.

Before you start

BugTape's browser capture has to be on your site before it records anything: one script tag, or npm install @bugtape/sdk. It does not catch every error, only the supported failures listed above. BugTape is hosted only. Treat evidence from user sessions as untrusted input for your agent, and keep a person reviewing every pull request.

Connect BugTape

npx -y bugtape connect cursor          # or claude, codex; click Allow once

BugHerd's MCP and BugTape's MCP can sit side by side in the same client. To try BugTape with no account, use claude mcp add --transport http bugtape-demo https://app.bugtape.ai/demo/mcp (read-only sample project). Setup guides: Cursor · Claude Code · MCP install.

FAQ

Is BugTape a BugHerd alternative?

For bugs nobody reports, yes. For client feedback pinned to page elements, no. BugTape has no pins, no guest review and no kanban board.

Does BugHerd have session replay?

BugHerd's features and pricing pages do not list its own session replay. Its help article says that when LogRocket is recording on the page, BugHerd saves a link to that LogRocket session on the feedback. BugTape browser issues can include a replay rebuilt from DOM events. A report can lack the events needed to rebuild one.

Does BugHerd capture errors automatically?

BugHerd's pages describe feedback that a person leaves, with a screenshot and browser details. Its public feedback widget lets site visitors report issues, but each report starts with a person. BugHerd's pages do not describe automatic error capture. BugTape files a report on its own for uncaught errors, failed requests and hung requests.

Can clients use BugTape without an account?

Your users do not need an account. The capture runs on your site, and they can press the optional bug button. There is no guest review flow for clients, so use BugHerd for client sign-off.

Can my coding agent read both?

Yes. BugHerd has an MCP server with OAuth sign-in, and so does BugTape. Add both to Cursor, Claude Code or Codex.

Does BugTape record video or screenshots?

No. Browser replay is rebuilt from DOM events. Screenshots and voice notes in the report form are not available yet. DOM replay vs screen video explains the difference for a coding agent.

Is BugTape free to try?

Yes. The read-only demo MCP needs no account. The Free plan needs no card and includes 150 credits a month.