TestMu AI’s Kane CLI now designs tests straight from the PRD

TestMu AI (formerly LambdaTest) launched the Assurance Lifecycle in Kane CLI — new commands that ingest requirement documents (PRDs, specs, policy pages), design acceptance tests permanently linked to those requirements, measure coverage from sealed evidence packs, and flag stale tests when a source document changes.

SOFTWARE desk — same-day primary for agentic-era testing that ties automated browser tests back to written requirements with auditable evidence, not just pass counts.

Background, still company: Kane CLI launched in April 2026 as a terminal-native browser automation tool for developers and AI coding agents, including Claude Code, Codex CLI, Cursor, and Gemini CLI. Teams describe a flow in plain language instead of picking page elements by hand; Kane CLI returns a pass-or-fail verdict with a step trace and screenshot, can export the flow to Playwright code, and can run headlessly in CI. Playwright is a common browser-test toolkit. CI means continuous integration — the automated checks that run when code is pushed. File that April launch picture as TestMu AI’s. This desk did not watch a session.

The job the company says most suites still miss: they report how many tests passed, not which written requirements were verified. Tests drift from the specs that inspired them. Coverage becomes a guess. When a requirement changes, nobody knows which tests went stale. File that gap as TestMu AI’s. This desk did not audit a customer suite.

Assurance flow, as the same wire has it, in seven stages: Capture — snapshot a PRD, spec, or policy page into a local store. Extract — an agent proposes use cases and cites the lines it read. Review — people promote, edit, or reject those proposals; agent output starts as derived, not silently trusted. Design — one acceptance test per scenario, tagged to the criteria it checks; design output is reviewed the same way. Execute — designed tests run and each run seals an evidence pack. Measure — coverage reports what a run proved and what the design still owes; missing coverage is a ranked gap, not silence. Maintain — when a source document changes, work extracted from the old snapshot is marked stale until it is checked again. A PRD is a product requirements document — the written “what the product must do.” Sealed evidence is a saved proof pack of what a test run actually did. File that seven-stage picture as TestMu AI’s. Exact command names stay in Sources. This desk did not ingest a PRD.

Store, still company: the assurance store lives in the project directory, is append-only with a replayable history, and stays local — sources, use-cases, designs, and review verdicts sit on the developer’s disk, not on a server. The company says the full lifecycle can run headless in CI or from an AI agent. File that local, append-only picture as TestMu AI’s. This desk did not inspect a store.

Availability, company: the Assurance Lifecycle is available today. Some extract, design, and reconcile steps run on the KaneAI agent and consume credits; review, coverage, and store operations are local and free. File that available-today / credits-vs-local split as TestMu AI’s. Version pins and install commands stay in Sources. This desk did not upgrade a CLI.

Named voice on the release: Mudit Singh, co-founder and head of growth, TestMu AI. He frames “validation debt” as agents writing code faster than teams can verify it. File the name, title, and that debt line as his, via TestMu AI. His quotes are color only and stay in Sources.

Plain English for the rest of the card: PRD = product requirements document, the written “what the product must do.” sealed evidence = a saved proof pack of what a test run actually did. assurance here = prove the written requirement, not just collect green checkmarks. acceptance test = a check that a specific promised behavior is true. stale = still in the suite after the source document moved on. Kane CLI = TestMu AI’s terminal tool that drives a real browser from plain-language intent.

PRIMARY here: TestMu AI’s 15 Sep 2026 PR Newswire company release — Tier A PRIMARY company source, the original record. The Kane CLI product page and assurance docs are product-home context, not a second originating newsroom. The Assurance Lifecycle introduce, the April 2026 Kane CLI launch picture, the seven-stage capture / extract / review / design / execute / measure / maintain flow, the local append-only store, available-today plus KaneAI-credits vs local-and-free split, and the Singh validation-debt line are company-attributed. Coverage and evidence claims stay company-attributed — not independently verified here. NOT claimed: customer logos, ARR, a valuation, that Kane CLI is open-source, independently verified coverage or evidence-pack results, that this desk tested Kane CLI or ingested a PRD, a stock tip, or investment advice. Distinct from the already-filed cyberhill-cerebro-llms, g5-labs-14m-seed, typesafe-ai-40m-seed, wso2-agent-manager-ga, cockroach-continuum, gensyn-open-1b-auditable, and creatify-boreal-launch.