4 posts tagged with "QA"

The traditional QA bottleneck is a well-known pain point for modern development teams. For years, the industry has pushed to "shift-left" – to move testing earlier in the development lifecycle. However, a major technical hurdle has always remained: the environment gap.

When QA happens on a global "staging" environment or only after code merges to the main branch, the feedback loop is too slow. Bugs found post-merge cause expensive context-switching for developers and delay releases.

Today, we’re bridging that gap. We’ve added full branch awareness to the TestChimp platform, enabling true shift-left testing at the PR level.

Why Branch-Aware Testing?​

Branch-aware testing means your QA process mirrors your Git workflow. Instead of testing "the app," you test the "feature-in-progress."

1. Test Authoring at the Feature Level​

You can now switch between repository feature branches directly within TestChimp. File versions are maintained per branch, allowing QAs to sync with branch-specific remote content.

Most importantly, QAs can author tests and raise Pull Requests from TestChimp that merge directly into the feature branch. This ensures that by the time a developer is ready to merge their code, the corresponding tests are already part of the PR.

[!TIP] Security & Outsourcing: Our new GitHub App-based approach means you don't need to give external QA resources full repository access. They can work exclusively on the tests and plans folders (with PRs raised via TestChimp platform), maintaining a tight security posture.

2. Branch-Specific Test Execution​

Gone are the days of manually pointing tests at different URLs. In your project settings, you can now configure a template string for branch-specific deployment URLs (e.g., Vercel or Netlify preview URLs).

When you run tests on a branch, TestChimp resolves the correct URL and injects it as a BASE_URL environment variable. Your scripts simply consume process.env.BASE_URL, ensuring they always target the correct preview deployment.

3. Exploratory Testing & Smart Bug Diffing​

Exploratory testing is no longer a "post-release" activity. All exploratory runs can now be executed against the branch-specific deployment.

Our agents are now smart enough to report only new bugs found on the feature branch compared to the default branch. This allows you to instantly see what UX, performance, accessibility, or internationalization issues were introduced by a specific PR – before they ever touch production.

4. QA Intelligence: Sliced by Branch​

In the Atlas page, you can now filter results by branch to see exactly how a specific screen or flow was affected by a PR. This level of granularity allows teams to answer the questions that actually matter during code review:

• "What user stories are breaking in this PR?"
• "Are unrelated scenarios being affected by these changes?"

Seamless CI Integration​

If you already have a CI pipeline that generates preview URLs, TestChimp fits right in. Simply pass that preview URL as the BASE_URL environment variable in your CI action, and your tests will execute against the live branch deployment with zero extra configuration.

Strategic Planning, Tactical Execution​

While test authoring and execution are now branch-aware, we’ve intentionally kept Test Planning artifacts product-scoped.

Strategy should be stable. Planning artifacts continue to sync with the repo's default branch, ensuring your high-level test coverage goals remain consistent even as individual features are developed and tested in parallel branches.

The Future is Shift-Left​

By moving QA participation closer to the development phase, you’re not just catching bugs – you’re preventing them from ever reaching the main branch. Branch-aware testing turns QA from a gatekeeper into a core part of the feature development engine.

• Which bugs are slowing down your checkout flow?
• Are there localization issues in onboarding causing drop-off?

Most QA setups fall short of answering such questions that matter to the business.

The Gap Between QA and Business Impact​

Traditional QA workflows surface issues --- broken buttons, layout glitches, validation errors, performance slowdowns. But when leadership asks:

• How is this affecting conversion?
• Is onboarding friction hurting activation?
• Which workflow is most at risk?

The connection between bugs and business outcomes is often missed entirely - resulting in QA being perceived as a cost center, than what it really is - revenue protection.

That's the gap we're addressing today by adding UX bug traceability to our QA intelligence layer.

Guided Exploration, Not Random Crawling​

Our secret sauce: Our exploratory agents don't wander randomly.

They use your existing automation tests as guides. Because tests are already linked to scenarios (via structured code comments), that same traceability naturally extends to exploratory runs and their findings.

This means every discovered issue can be traced back to:

• The scenario
• The user story
• The business objective behind it

Exploration becomes business-contextual --- not isolated.

Structured for Insight at Any Level​

In TestChimp, user stories are organized in nested folders. That structure becomes powerful when paired with traceable exploratory results.

Insights roll up automatically:

• By area of the application
• By workflow
• By product surface
• By team ownership

You can zoom in to a single scenario or zoom out to understand impact across an entire product area.

Beyond a Floating Bug List​

Instead of maintaining a detached list of issues, you gain visibility into:

• Which flows are degrading user experience
• Which exact bugs are causing latency in key user journeys - that mater to your revenue
• Where UX friction is tied to user retention

It's no longer just about "bugs found."

It's about translating them understanding what is hurting your revenue --- and prioritizing accordingly.

We used to live in forms.

Historically, dropdowns and text fields were the default way we planned and managed work. But in the agentic era, the winning UX isn’t a fancy form. It’s plain, boring text.

We already see it everywhere. We use skills.md for upskilling agents. claude.md for context. Spec-based development in Cursor. But look at your test planning tools. Jira, Linear—they were built in a pre-AI era. They’re database-centric, form-heavy, and fundamentally hostile to agentic workflows.

Shouldn’t testing be as modern as coding?

That’s why we’ve reimagined test planning for the agentic era. We call it Test Planning as Code.

Plans as strongly typed markdown​

In TestChimp, your plans live as strongly typed markdown files—user stories and test scenarios as .md files with YAML frontmatter, organized in folders and version-controlled alongside your codebase.

There are some pretty significant advantages to maintaining stories and scenarios as simple .md files.

First, they sync to your code repository. That means your coding and testing agents can read them and work on them directly. No proprietary API, no “export for AI”—just the same files your team already uses.

Second, you can organize them in a nested folder structure however you want. By area. By journey. By team. That structure gives agents broader context. They see related stories and linked scenarios, not just isolated tickets floating in a database.

This is what actually gets stored in your repo. No proprietary formats. No lock-in. Just plain markdown.

“I don’t want to manage status in a text editor”​

Humans still need workflows. We need priorities, due dates, and assignees.

TestChimp layers those workflows on top of the files. You get the familiarity of a structured UI for human workflows—rich forms, status dropdowns, filters—without losing the benefits of file-first planning. The source of truth stays in the files; the platform makes them easier to work with.

And because TestChimp indexes everything, the AI can actually work with your plan. It can help write or refine a user story more accurately. It can suggest relevant test scenarios. It can even detail them out, grounded in your actual requirements.

Linking tests is trivial​

Once you have scenarios, linking tests is trivial. Just add a comment in your test code:

// @Scenario: Login - Invalid Credentials Error

No spreadsheets. No manual mapping. No juggling multiple tools.

Export to Git keeps your test plan in the repo—stories and scenarios live under a path you choose, with full history and pull-request workflows. Your agents and your CI see the same files.

Coverage at any granularity​

As tests run, coverage insights update automatically. And because your stories are organized in folders, you can see coverage at any granularity—per story, per area, per component.

If you’re working in a team where different groups own different parts of the system, you already know how useful this is.

You can finally answer the question: Which scenarios are due next week, ready for testing, but still missing coverage? No spreadsheet. No manual roll-up. Just select the folder, apply the filters, and look at the Insights tab.

Wrapping up​

Test Planning as Code is a different take on where test artifacts should live and who should be able to use them. Files in the repo instead of rows in a database; workflows layered on for humans, and that same file structure giving agents the context they need. If that approach resonates—or you’re just curious how it works in practice—we’ve documented the full workflow in the Test Planning section: authoring user stories, authoring test scenarios, export to Git, and requirement traceability.

Q: How do you currently get requirement traceability?
Which user stories and scenarios are covered by tests, and what’s failing?

A:
For most teams, it looks something like this:

User stories live in Jira.
Test cases live somewhere else.
The mapping between them lives in an Excel sheet that someone manually maintains.

That spreadsheet is periodically uploaded to a test management tool like PractiTest. Then test execution results are pushed via an API to get a view of coverage and failures.

It works—until it doesn’t.

The problem with today’s approach​

This is how requirement traceability is typically achieved today: a hodgepodge of tools stitched together with process and hope.

• Multiple sources of truth
• Manually maintained spreadsheets that inevitably go stale
• Fragile workflows that break as teams and test suites scale

No single system actually owns the full picture. Instead, teams spend time keeping artifacts in sync rather than improving product quality.

A simpler model with TestChimp​

In TestChimp, requirement traceability isn’t an afterthought. It’s built in.

You already author detailed user stories and break them down into meaningful test scenarios directly in the platform - with AI assistance that understands your product through your existing test scripts and documentation.

Linking tests to those scenarios is intentionally simple. In your test script, add a comment:

// @Scenario: <scenario title>

That’s it! TestChimp takes care of the rest:

• Automatically links tests to scenarios
• Tracks execution results across runs
• Aggregates outcomes at scenario, story, and suite level

You get clear, real-time dashboards that let you answer business relevant questions:

• Which user stories are missing test coverage?
• Which scenarios are currently failing?
• Which tests are flaky or unstable?

All without juggling multiple tools or maintaining brittle Excel sheets.

⸻

One system, end to end​

Instead of retrofitting traceability after the fact, TestChimp treats it as a first-class concept - connecting requirements, scenarios, and executions in one place.

• No spreadsheets.
• No manual syncing.
• Just a single system that understands what you’re building, how it’s tested, and where the gaps are.