Skip to main content

Changelog

What each TestChimp platform release shipped, newest first. A machine-readable copy is at /changelog.json.

1.0.17 - 2026-10-14​

Highlights​

Git sync now handles GitHub and GitLab rate limits gracefully: when a busy repository hits its API limit, TestChimp pauses and finishes the sync automatically after the limit resets instead of silently skipping files.

Bug fixes​

  • GitHub and GitLab integration: Large or frequent pushes no longer leave tests and plans partially synced when the Git provider's API rate limit is reached. Sync waits for the limit to reset and then completes, and a repository connected to several projects downloads each changed file only once per push, so busy repositories stay well within their limits. Connecting a repository or creating a project is never held up by a pause from background sync.

1.0.16 - 2026-10-14​

Highlights​

GitHub sync is more dependable: large or bursty pushes are now processed reliably in the background instead of occasionally being dropped, and new users can watch a short intro video right from the dashboard.

Improvements​

  • Getting started: The "Learn More" link on the How TestChimp Works tip now plays the TestChimp intro video in-app instead of sending you to the docs.

Bug fixes​

  • GitHub integration: Pushes that touch many tests, plans, or a newly created branch now sync reliably. Heavy updates are processed in the background, and simultaneous pushes to a new branch no longer fail with a conflict.

1.0.15 - 2026-10-14​

Highlights​

ExploreChimp now runs on your machine through your coding agent instead of TestChimp cloud, and joining a team from an invite is smoother—especially when you are re-invited or already signed in in the browser.

Improvements​

  • ExploreChimp: Run ExploreChimp locally via your coding agent. The app gives you a ready-to-copy prompt from SmartTests and onboarding; cloud-hosted exploration wizards and the older UI Test Studio and Test Tree Map surfaces are retired.
  • Team invitations: Accepting an email invite keeps the verification from the invite link, so you are not asked to verify again. Re-inviting a teammate who was removed earlier sends them a fresh invite.

Bug fixes​

  • Team invitations: Opening an invite or signup link while a stale browser session is still signed in no longer fails with a server error.

1.0.14 - 2026-10-10​

Highlights​

TestChimp is now more resilient: services that stall recover on their own within minutes, and merging pull requests in large, many-branch projects no longer puts the platform at risk of going offline.

Improvements​

  • Reliability: stalled platform services are now detected and restarted automatically, keeping outages short.
  • Team invitations: invite emails now include a link to set up the TestChimp QA bot on Grok Bot.

Bug fixes​

  • Git integration: merging a GitHub pull request or GitLab merge request in a project with many branches and test files no longer causes runaway memory use that could make TestChimp unavailable.

1.0.13 - 2026-10-07​

Highlights​

The automation batch list now honors your platform filter, and QA Bots have a dedicated documentation section covering how the bot swarm works and how each role on your team can put it to use.

Improvements​

  • Documentation: a new QA Bots section explains how a swarm of QA bots works alongside your team, with a how-to guide for installing and using bots as a team lead, developer, or QA engineer.

Bug fixes​

  • Executions: changing the platform selection in the automation batch list now updates the results immediately and shows only batches that ran on that platform.

1.0.12 - 2026-10-07​

Highlights​

ExploreChimp UX analysis works again for local and agent-driven Playwright runs: screen states marked during your tests are analyzed and their bugs are surfaced again.

Bug fixes​

  • ExploreChimp: screen states marked with markScreenState during local and agent-run Playwright tests were rejected before analysis. Explorations still completed, but no screens were analyzed, no credits were used, and no UX bugs were reported. They are now analyzed and charged credits as normal, with no client update needed.

1.0.11 - 2026-10-06​

Highlights​

A polish release: handing a story or scenario from the Plans tab to ChimpHands in TestChimp Studio now reliably opens the plan file, and the Grok QA bot setup now links straight to the published bot.

Bug fixes​

  • Plans: sending a story or scenario to ChimpHands from the Plans tab now opens the plan file in Studio's editor. Previously, the file could fail to open when it wasn't on your machine yet or when the plans folder path was doubled. Local edits to existing files are never overwritten.
  • QA bots: the Add bot screen's Grok option now links to the published TestChimp QA bot instead of a placeholder page.

1.0.10 - 2026-10-06​

Highlights​

Every workflow execution now has a clear owner. An assignee and an optional CC list show who needs to review or unblock a run, and the assignee is notified in the app, by email, or in Slack, whichever they prefer.

New capabilities​

  • Workflows: each workflow execution now has an assignee and an optional CC list. When a run needs plan approval, TestChimp assigns it automatically to the person who started it, if there is one. A new pane under the execution title shows the assignee and CCs as profile pictures, and an Update button lets you reassign the run or change who is CC'd. The executions list has a new Assignee column.
  • Notifications: a new Workflow execution assigned notification lets you know when a run is assigned to you. It's on by default, and you can choose whether it also reaches you by email or Slack in your notification settings. QA bots can subscribe to the same event.
  • AI agents: coding agents and QA bots connected through TestChimp's MCP server can now invite teammates to a project and update a workflow execution's assignee and CC list.

1.0.9 - 2026-10-05​

Highlights​

TestChimp's hosted MCP server now works with sign-in from popular AI coding agents and IDEs (Cursor, Claude Code, Codex, Windsurf, Kiro, and others). You can connect without pasting keys into config files, and each QA bot can now work on its own project.

New capabilities​

  • AI agents: connect to TestChimp's hosted MCP server by signing in from your agent or IDE, including apps that return to the agent after sign-in (such as cursor://). Once connected, an agent can fetch your project's credentials to set up local test runs (Playwright reporter and CLI), so it doesn't need a hand-edited config. Existing manual MCP setups keep working unchanged.
  • QA bots: each bot now works on its own project, even when several bots share one TestChimp connection. You connect once, and each bot uses its own project API key. The Add bot and sign-in screens explain this setup.

Improvements​

  • AI agents: sign-in sessions stay active when an agent refreshes its connection from several requests at once. Previously this could sign you out.

1.0.8 - 2026-10-04​

Highlights​

Every team member can now connect their own QA bot (Grok Bot) to a TestChimp project. It hears about their pushes, assignments, releases, and test runs, proposes the next QA step, and acts only after they approve.

New capabilities​

  • QA bots: a new My Bots tab in user settings lets you connect a Grok Bot to any of your projects in a single sign-in step. Pick which events it should hear about, verify its webhook, and remove a bot you no longer need. The bot works with your existing permissions. It can update stories and scenarios from your pushes and meetings, write E2E tests, plan issue fixes, triage failing test batches, and give release go / no-go heads-ups, each after you approve.
  • QA bots on your computer: requirement updates from AgentWatch, folder mapping, and local test runs run on your own machine through the TestChimp CLI. TestChimp Studio doesn't need to be installed.
  • AI agents: coding agents and bots connected through TestChimp's MCP server can now send feedback or report problems straight to the TestChimp team.

Improvements​

  • Team management: inviting a teammate past your plan's seat limit now explains the limit for your plan (free trial, Indie, or Teams) and offers an upgrade, instead of showing a generic error. Growth plans have unlimited seats and never see this prompt.

Bug fixes​

  • API coverage: OpenAPI specs larger than 1 MB now sync from GitHub and GitLab (up to 20 MB). Previously they were skipped, which left API coverage empty.

1.0.7 - 2026-10-02​

Highlights​

Branch search now finds every branch that matches what you type, and picking folders, spec files, and branches is tidier and stays fully on screen, even on smaller displays.

Improvements​

  • Git integrations: mapping a tests or plans folder now opens in a centered dialog that fits smaller screens. The mapped plans folder shows the actual name of your default branch instead of "repository default".
  • API coverage: each OpenAPI spec now shows its branch and file on a single row, with your default branch's real name. Long names are shortened, and you can hover to see the full name. Edit and remove are icon buttons at the end of the row. The file picker opens beside the button and is more compact, so it no longer runs off the screen.

Bug fixes​

  • Branch selection: searching for a branch by name now returns every match, including your own branches whose names start with "testchimp".

1.0.6 - 2026-10-02​

Highlights​

When you map your plans folder in the GitHub or GitLab integration, you can now pick the branch to sync with in the same step, and the branch list loads quickly even on very large repositories.

Improvements​

  • Git integrations: the plans folder mapping dialog now asks which branch to sync with, and the folder browser shows that branch's contents. It starts on the default branch or the branch you already configured, so existing setups are unchanged. Mapped plans folders show their branch next to the folder. You can switch branches from the same dialog without remapping the folder.
  • Branch selection: branch pickers no longer time out on repositories with many branches. They show the default branch, your configured branch, and branches active in the last 30 days, and you can search by name for any other branch.

1.0.5 - 2026-10-02​

Highlights​

Teams can now choose which Git branch TestChimp syncs their plans with, and which branch each OpenAPI spec is read from, instead of always using the repository's default branch. Git sync also stops labelling new plans that are waiting in an unmerged pull request as "deleted in remote".

New capabilities​

  • Git integrations: choose the branch your plans folder syncs with in the GitHub and GitLab integration settings, or through MCP. It defaults to the repository's default branch, so nothing changes until you pick one. Switching branches re-aligns the sync so it doesn't fill up with false changes, and if the chosen branch is later deleted, sync falls back to the default branch and tells you.
  • API coverage: when you add OpenAPI spec files on the API page, you can pick the branch each spec is read from. TestChimp remembers the branch with the file path and reads updates from it. It defaults to the repository's default branch.

Bug fixes​

  • Git sync: plan changes pushed to other branches, such as TestChimp's own sync pull-request branches or your feature branches, no longer count as synced to your plans branch. New plans waiting in an unmerged sync pull request are no longer shown as "deleted in remote", and deleting one in TestChimp is no longer hidden from outgoing changes.
  • Sign-in errors: requests made with an invalid personal access token, for example from TestChimp Studio or the Chrome extension, now get a clear "unauthorized" response instead of a server error.

1.0.4 - 2026-10-01​

Highlights​

AI agents can now search and browse your team's meetings through the TestChimp MCP and API. They get the same date, label, participant, and full-text filters as the Meetings page. Git sync also opens straight to results instead of hanging.

New capabilities​

  • Meetings for agents: agents connected through MCP or the API can list team-wide meetings, newest first. They can filter by date range, labels, participants, or participant email domains, and run full-text search over titles and transcripts, with highlighted snippets. A companion call returns the labels, participants, and domains available to filter on.

Bug fixes​

  • Git sync: starting a sync no longer leaves the page stuck loading until you refresh it. Analysis starts as soon as the project is ready, for both test and plan sync, and the page shows a clear error if it can't work out which branch to sync against.

1.0.3 - 2026-09-29​

Highlights​

The project Overview now reports accurate automation numbers for teams working across many branches, and TestChimp points Studio users to AgentWatch so their QA plan keeps pace with their coding agents.

New capabilities​

  • Action tips: users who haven't turned on AgentWatch in TestChimp Studio see a tip, on web and in Studio, that explains how it keeps user stories and test scenarios in sync with coding-agent sessions and links to the AgentWatch docs. The tip goes away once AgentWatch is enabled.

Improvements​

  • Project overview: the automation widgets load faster on projects with hundreds of branches, and the summary and charts on one page share the same data.

Bug fixes​

  • Project overview: the automation status widget counts each test once across all branches instead of showing only the tests on a single branch. A test whose suite was renamed on one branch is still counted once.

1.0.2 - 2026-09-28​

Highlights​

Meet Bots now do real work after the call. Every meeting gets an AI summary, and bots can read your stories, scenarios, and connected code to update plans or start authoring workflows. A new meetings workspace adds full-text search, filters, and per-meeting privacy controls.

New capabilities​

  • Meet Bots: each finished meeting gets an AI-generated summary, shown in a new Summary tab next to the transcript.
  • Meet Bots: bots work through multi-step tasks during and after meetings. They can read and update stories and scenarios, search related plans, browse your connected git repository, and start plan-authoring workflows from meeting decisions. While a bot works, it posts progress updates in the meeting chat.
  • Meetings: search the full text of titles and transcripts, and filter meetings by date range, participant, or participant email domain.
  • Meetings: control who can see each meeting: participants only, all team members, or participants plus specific teammates. Participants are recorded from calendar invites, so people who joined from a shared meeting room are included.
  • Meetings: save a filtered set of meetings as a meeting set and hand the whole set to agents for follow-up work.

Improvements​

  • Releases: creating or updating a release returns immediately, and release deltas are computed in the background.
  • Releases: the release list hides archived releases.
  • Navigation: Documentation moved into the settings menu to keep the sidebar uncluttered.
  • Getting started: the "Setup local agent" guide has a TestChimp Studio tab with a one-click install. No configuration is needed.

Bug fixes​

  • Workflows: executions whose plan was already approved, or that were started by AgentWatch, now close as completed instead of staying on "Plan ready" with Accept and Reject buttons still showing.

1.0.1 - 2026-09-27​

Highlights​

Meet Bots bring TestChimp into your Google Meet, Zoom, Microsoft Teams, and Webex calls to capture tribal knowledge and turn meeting decisions into QA workflows, while AgentWatch keeps stories and scenarios current by learning from your team's Cursor and Claude Code sessions.

New capabilities​

  • Meet Bots: send TestChimp bots to Google Meet, Zoom, Microsoft Teams, and Webex calls, or schedule them from a connected Google or Outlook calendar. Transcripts land in your project as labelled, organized meeting notes.
  • Meet Bots: mention the bot during a meeting to get answers grounded in your project's stories, scenarios, and test results.
  • Meet Bots: define custom bots with their own policies alongside the built-in QA and Sales bots. Meeting decisions become searchable context for planning and QA workflows.
  • AgentWatch (with TestChimp Studio): watches your local Cursor, Claude Code, and other agent sessions, spots when business rules or UX expectations get refined, and drafts a plan to update your stories and scenarios. Runs on every git push or on a schedule, with an Agent Sessions Explorer to browse, search, and re-run extractions.

Improvements​

  • Planning overview: scenarios are split into manual and automation candidates, so automation coverage insights only count scenarios meant to be automated.
  • Test runs: the test run viewer shows clearer scope tags for what each run covers.
  • Getting started: the "How TestChimp Works" tip links straight to the TestChimp Studio download, and a new tip nudges teams to record their first meeting.

Bug fixes​

  • Manual testing: notes on manual steps without a screenshot now show on a placeholder instead of having nothing to anchor to.
  • Reliability: fixed out-of-memory crashes when ingesting results and resolving coverage for large test suites.

1.0.0 - 2026-09-27​

Highlights​

TestChimp is the AI-native QA workflow layer for fast-moving teams: plans, tests, and quality signals live in your repo, and agents keep them current on every PR.

New capabilities​

  • Agent-orchestrated QA: the /testchimp skill, TestChimp Studio, Automations, and ChimpHands cloud agents plan, write, run, and repair tests on every change.
  • Tests you own: Playwright and Mobilewright SmartTests in your repo, traced to markdown requirements. No vendor lock-in.
  • Performance testing: k6 journeys linked to real scenarios, with baselines and regression comparisons.
  • Coverage intelligence: TrueCoverage matches real-user traffic to your tests, and API contract coverage shows the gaps that matter.
  • Exploratory UX analysis: ExploreChimp walks your app to find UX bugs and route them into issue triage.
  • Release governance: release checks bring manual and automated runs, DAST, and dependency scans into one go/no-go view.