fdbck.

Changelog

What's new in fdbck, newest first.

v1.12.0

Every error comes with its fix

  • Errors now show the button that fixes them — Run the fix again, Try again, or a link to the setting — instead of a sentence telling you what to go do.
  • A fix that fails after you approve its plan now says so, with Retry, instead of waiting for a review that will never come.
  • The worker checks a fix against the exact code it changed, so a stale local checkout no longer makes good fixes fail.
  • When a repo's own checks fail without the change too, the pull request says the checks can't judge this fix, rather than blaming it.
  • Style-only changes are never marked Verified: the pull request names the page to look at before you merge.
v1.11.0

Repair pull requests without leaving fdbck

  • Resolve: when a fix's pull request conflicts with your main branch, one tap has the worker redo it on the latest code.
  • Fix checks: when CI fails, the worker reruns the fix with the failing checks in hand.
  • CI status counts every check on the pull request, so one green job can't hide a red one.
  • Pull request state is re-read when you open an item, so Resolve appears as soon as your main branch moves.
  • Starting an action hides the others until it finishes, so nothing can be started twice.
  • Mark a report as Ignore similar to keep the worker from acting on look-alikes.
  • Rename your board's Planned, In progress and Shipped to your team's own words — on the roadmap, every card, item pages and share images.
v1.10.0

Start without connecting GitHub

  • Connect later: create a project, embed the widget and get AI triage and a public board before connecting a repo. Attach GitHub whenever you're ready.
  • Link related reports from the item menu. Links show on your public board when both items are published, and the worker reads them for context.
  • Plans read your code before they're written, make the design decisions themselves, and ask only what they can't responsibly decide.
  • Plans can only name files that exist, and they list every copy of a page they change.
  • A plan being written shows live progress instead of a bare spinner.
  • A worker result is never lost to a dropped connection — reports retry until they land.
v1.9.0

Built for phones

  • On phones, Send feedback lives in the menu and Report this appears beside errors — no floating button covering your app.
  • The feedback form opens as a sheet that sits above your own app's layers.
  • The installed app picks up a new version when you come back to it.
  • Worker status now sits in the mobile menu's title row.
  • Empty states and every page clear the phone's status bar and don't scroll.
  • Pull requests now say exactly what was checked: Verified, Checks pass, or Fix not exercised by tests.
  • Fixed: Project settings and public boards could fail to load.
v1.8.0

Pro, and sign-in that just works

  • Upgrade to Pro from your account with Stripe checkout, and manage billing in the Stripe portal.
  • Sign-in links survive corporate email scanners. Opening the link no longer signs anyone in until you confirm.
  • Waiting on your GitHub organization's owner to approve fdbck? We email you the moment they do.
  • The pull request menu on phones opens upward and stays on screen; the mobile action bar has one menu instead of two.
v1.7.0

Run your pull requests from fdbck

  • Merge, mark ready, close, re-run CI and request changes on a fix's pull request without opening GitHub.
  • GitHub issues sync in as feedback, labelled and triaged like any other report.
  • Merging duplicates shows what overlaps, with one-click undo.
  • Add feedback directly from your inbox.
  • Shipped items lock their vote count on the public board.
  • Runs show whether they're queued, paused, blocked or stopped, instead of one generic state.
  • Choose whether an item's timeline reads oldest or newest first.
  • The widget captures page context in React Native and Flutter apps.
  • Public board items get their own share images.
  • Every control you tap on a phone meets the 44px touch size.
  • A fix can no longer pass its checks by rewriting the test that says it's wrong.
v1.6.0

Every item reads as a pipeline

  • Each item shows its journey — Reported, Triaged, Planned, Fixed, Shipped — with what's happening and what's waiting on you at each stage.
  • Approve right where the decision is made: on the triage you're reading, or on the plan.
  • Written plans get their own Approve plan button, and approving starts the fix.
  • Settings shows whether planning first actually leads to cleaner merges on your project.
  • On phones, the stages stack without nested boxes; filter the list to just what needs review.
v1.5.0

Public boards, your way

  • Put your public board on your own domain.
  • Search the board, reply to items as the owner, and page through long boards with honest ranking.
  • New boards start with example items that show the whole loop, without spending your agent's time.
  • Six-step onboarding ends with a live board and a running worker.
  • Readable item links: /board/your-product/item/the-title.
  • Automation rules: when an item is triaged, gets enough votes, ships or fails CI, open a pull request, write a plan, tell voters or stop and ask you — with built-in limits so a rule can't run away.
  • fdbck's own public board lives at fdbck.app/board.
v1.4.0

Bring your backlog

  • Import feedback from a CSV export of another feedback tool, with the columns mapped for you.
  • Live intake from Zendesk, Intercom and Slack.
  • Plan first: for larger requests the worker writes a short plan you approve before any code is written.
  • Comment on any item. Your comments guide every later run.
  • Duplicate reports count as demand for the original, not noise.
  • Hitting your AI plan's usage limit pauses a run and resumes it automatically, instead of marking it failed.
  • When a reporter attaches a screenshot, the worker looks at it while it fixes.
  • Run more than one worker and give each a job: triage on one machine, fixes on another.
  • The widget records what the user did before reporting, not just what broke.
v1.3.0

Fixes that check themselves

  • The worker runs your repo's own typecheck, lint and tests on every fix by default, and gets a chance to correct its own failures before you see a pull request.
  • Each fix gets a working copy of your project with its dependencies and settings in place, so checks reflect the change, not a broken setup.
  • Triage reads a map of your repo, so it classifies with the code in view.
  • Per-repo instructions for the agent, written once in Settings.
  • Install the worker as a background service with one command.
  • CI failures caused by billing or infrastructure are told apart from failures in the code, so the agent isn't blamed for either.
  • Step to the previous or next item from an item's page.
  • The widget keeps your draft when you click outside it.
v1.2.0

Polish from daily use

  • Settings has its own navigation, and the public board moved into a Widget & Board tab.
  • The public board is light by default, with a theme toggle and a ?theme= option for embeds.
  • Notes you leave on an item feed straight into triage, and Needs review always has a way forward.
  • The worker tells you when an update is available, updates itself, and has --verbose and --help.
  • Model menus stay current with what each agent supports.
  • Twenty-three fixes found by running fdbck's own feedback through fdbck.
v1.1.0

Teams, and a board people can vote on

  • Teams: everyone with access to your GitHub repo gets the matching access in fdbck — no invites to manage.
  • Sort the public board by Trending, Top or New, with clearer cards.
  • Build your own board UI on the Board API: read items, vote and count votes from any site.
  • Everyone who upvoted an item hears when it ships.
  • Each public item has its own page, and every board has a changelog with an RSS feed.
  • Press ⌘K anywhere to jump to a page, switch project or search your feedback.
v1.0.0

Close the loop — everywhere

  • Notification channels: pipe new feedback, opened PRs, merged fixes, and failures to Slack, Discord, Telegram, or any webhook (HMAC-signed, so Zapier/n8n/your own service work too). Configure up to 10 in Account → Notifications, each with per-event toggles and a Send-test button.
  • Reporters can now leave an email in the widget — and get exactly one “Your feedback shipped” email when the fix merges, with a one-click opt-out. Never shown publicly, never sent to the AI agent.
  • fdbck speaks MCP: connect your inbox to Claude Code or Cursor and let your coding agent list, inspect, approve, and retry feedback as tools — signed in with your worker's access key. See the new MCP docs.
  • The agent learns from you: when you merge one of its PRs with edits, fdbck captures what you changed and feeds it into future runs on that project. A new confidence-calibration strip in the Track Record card shows whether high-confidence items really merge clean.
  • Live run checkpoints: an in-flight fix now shows exactly what the agent is doing (“editing src/billing.tsx · 12 reads · 3 edits”) instead of a spinner — in the dashboard and via MCP.
  • Pricing is public: flat, cheap, and unlimited where it counts — AI triage and fixes are never metered, on any tier. Free covers your first project; Pro is a flat $19/mo ($15/mo yearly).
v0.9.0

A premium feedback experience for your users

  • Configure what users see after they submit — a custom confirmation message, footer links, an optional redirect, and a “Powered by fdbck” toggle — all from Settings → Widget messaging, no code changes.
  • Every submitter gets a private “Track your feedback” status link: a clean, honest timeline from Received to Shipped, with no account needed.
  • New public feedback board: publish the items you choose to a curated, embeddable roadmap (Planned / In progress / Shipped) with anonymous upvotes. Nothing is public until you publish it.
  • Tougher on PII and abuse: smarter redaction (real card numbers, IPs, and emails go; your order ids and version numbers stay), a live character counter, and untrusted feedback is always treated as data — never instructions — when it reaches the agent.
  • Sharper triage: the agent now uses the page, component, and app version to classify better, and flags anything ambiguous or risky as “Needs review” so it waits for you instead of auto-opening a PR.
  • New docs: a direct Feedback API reference and a Public board guide.
v0.8.0

A more trustworthy, more polished worker

  • Verify before opening a PR: point the worker at your repo's tests/build (`"verify": true` in worker.json) and it runs them in the isolated worktree first. A fix that fails still opens a draft PR — clearly flagged — but is never auto-merged.
  • Auto-merge got stricter: it now refuses changes to CI, Dockerfiles, lockfiles, and shell scripts, and won't merge anything that failed verification.
  • Lost work is recovered automatically. If your laptop sleeps or the worker dies mid-job, a reaper requeues the job within minutes instead of leaving it stuck forever.
  • The dashboard now shows when the worker is actively Working (not just online), warns if a job looks stuck, and displays the worker version + machine name.
  • Hardened: your feedback text is handled as data the agent reads, never as instructions it could be tricked into following; the worker warns if your token file is readable by others.
  • The worker CLI itself got a glow-up — a proper banner, color-coded live status, and a readable run-by-run log as it triages and fixes.
v0.7.0

Daily summary email

  • Turn on a once-a-day digest in Account → Notifications: PRs your worker opened, anything that failed, and what's awaiting your review — so you can review on your own schedule without watching the dashboard.
  • Opt-in and off by default. Quiet days are skipped (no empty emails).
v0.6.0

Open the widget from any element

  • The floating button is now just the default. Add `data-fdbck` to any element — a branded button, a footer “Support” link, a menu item — and it opens the same feedback modal. No JavaScript, and it works for elements rendered after load.
  • `data-fdbck-type="bug|feature|other"` lets a trigger preselect the feedback type, so a “Report a bug” link lands users on the right form.
  • New JS API for framework apps: `fdbck.open()`, `fdbck.close()`, `fdbck.toggle()` — `open()` takes an optional `{ type }`.
  • Going launcher-less (great on mobile)? Set `data-hide-fab="true"` on the script tag to drop the floating button and drive the modal entirely from your own UI.
v0.5.0

Multi-vendor agents

  • The worker now runs whichever agent CLI you've installed — Claude Code, OpenAI Codex, or Google Gemini — not just Claude.
  • Pick the vendor and model per phase (triage vs. code) in Account → Models. Still zero keys stored on fdbck: it all runs on your own subscription.
  • Power users can override the exact command per vendor in worker.json via a `commands` map.
  • Every PR fdbck opens now carries a brand-orange `fdbck` label, so they're easy to filter in your repo's Pull requests.
v0.4.0

Onboarding & docs

  • First-run setup checklist in the inbox — embed the widget, run the worker, send a test — each step lights up green the moment it's done.
  • Live local-worker status: a pill in Account shows whether your worker is online, idle, or offline.
  • Public docs at /docs: getting started, the full widget embed reference, running the worker, automation, and troubleshooting.
  • “Send test feedback” fires a synthetic item through the whole pipeline so you can prove the loop without waiting for real traffic.
  • Auto-PR and Auto-merge now use plain-English presets (Conservative / Balanced / Aggressive) with the raw thresholds tucked under Advanced.
  • fdbck-worker doctor — one command validates your whole setup and tells you exactly what's missing.
  • Inline help tooltips and Docs links throughout settings; regenerate your widget key in one click.
v0.3.0

Zero server keys — your Claude, your machine

  • fdbck no longer stores any model keys. Triage and fixes both run on your own machine via the fdbck-worker CLI and your own Claude Code.
  • The local worker claims feedback, triages it, writes the fix in an isolated git worktree, and opens a draft PR — all with your own credentials.
  • Per-phase model preferences: pick a fast model for triage and a stronger one for code.
v0.1.0

The core loop

  • Embeddable feedback widget with automatic, PII-scrubbed page-context capture.
  • Triage that classifies type, severity, and confidence, then gates which items get an automatic fix.
  • Draft pull requests against an isolated branch, with optional auto-merge for high-confidence bug fixes.
  • Multi-app support, per-project target branch and monorepo root, and a trust dashboard tracking clean-merge rate.