Testsy descriptions and launch assets
Accurate Testsy product descriptions, official links, example reports and guidance for mentioning the product in a directory, newsletter or community.
A short description#
Testsy is an AI tester for small teams shipping SaaS products. Describe a user journey in plain English, review the plan, and inspect the result with browser evidence.
For a short listing: Plain-English release checks with evidence for small SaaS teams.
A longer description#
Testsy helps small teams check the customer journeys that matter after a release: signup and email confirmation, sign-in, creating and saving a record, or publishing a document. Describe what should happen, review the proposed steps and checks, and run the saved test against a reachable environment. Results show the verdict, observed values and available replay evidence so a developer can investigate a failure and rerun after a fix. Teams can also run saved tests through CI, the API or an MCP-compatible coding agent.
Official links and examples#
| Resource | Link |
|---|---|
| Product and current plans | Testsy home |
| Editable release smoke test | Try a release check |
| Editable signup and email test | Try signup verification |
| Product icon | Testsy SVG icon |
| Inspectable sample results | Public example reports |
| Setup and documentation | Documentation |
| Coding-agent connection | MCP guide |
| Release automation | GitHub Actions guide |
| Support and listing corrections | Contact Testsy |
The public reports use synthetic example data. A report demonstrates its particular test; it is not a claim of universal coverage or a result from your own product.
Integration downloads#
These versioned release packages connect an agent to the hosted Testsy service. Native packages contain setup instructions and the shared skills; they do not include an API key or the Testsy engine. Keep credentials in the client's secret mechanism. A download is not a public marketplace approval.
| Package | Download |
|---|---|
| Integration source, MIT license and build instructions | Source archive |
| Claude Code native package | Claude package |
| Claude Code local marketplace | Claude catalog |
| Codex skills package | Codex package |
| Codex local marketplace | Codex catalog |
| Cursor package for supported import or submission | Cursor package |
| Gemini CLI extension | Gemini extension |
| Portable skills with a separate MCP connection | Portable skills |
| OpenAI OAuth candidate, pending live qualification and directory review | OAuth candidate |
| GitHub Action with bundled CLI | Action package |
| MCP Registry submission metadata | Registry metadata |
SHA256SUMS and release metadata identify the published files. Download the checksum file beside the archive and compare its SHA-256 before installation. For example, after downloading the Claude package and checksum file into the same directory:
shasum -a 256 testsy-claude.tar.gz
# Compare the hash with testsy-claude.tar.gz in SHA256SUMS.
tar -xzf testsy-claude.tar.gz
claude --plugin-dir ./testsy-claudeSet QA_AI_API_KEY securely in the client environment before connecting the native API-key package. Read the package's README for your client; native packages do not need a package build step. The source archive includes the release README and build/check tools. Read MCP setup for workspace permissions, browser authorization and revocation.
The local catalogs are installable from disk. They are separate from approval in a hosted directory. The action package can be copied into an application repository; follow its README rather than inventing a remote uses: address. The OAuth candidate needs deployed-service and clean-client validation before it is advertised as an approved ChatGPT integration.
Describe the scope accurately#
A passing test covers the reviewed journey, environment and observations. A failed assertion remains a failure. Blocked, inconclusive and infrastructure-error results are distinct outcomes and must not be promoted as passes.
Do not describe Testsy as guaranteeing bug-free releases, autonomously approving deployments, fixing code without review, or reaching arbitrary private networks. Refer readers to the current app testing guide for supported build types and availability instead of promising physical-device coverage.
Use the current home page for trial and pricing details rather than copying plan numbers into permanent artwork. Directory listings and integrations may have different review or publication states; direct readers to the installation instructions that are actually available.
Demonstrate one complete journey#
A useful short demo shows the requirement, the reviewed plan, the run and the evidence for one check. For a SaaS example, create a synthetic task and show whether it survives a reload, or verify a new account through its confirmation email.
If a run fails or cannot establish the answer, keep that outcome visible and explain what it proves. Show a separate passing rerun after a deployed fix rather than replacing the earlier evidence.
Mention Testsy in a community#
Disclose your relationship to Testsy and follow the community's promotion rules. Adapt the description to the audience's actual question; avoid automated message sequences and requests for artificial votes or reviews.
Write a Show HN submission yourself in accordance with Hacker News's current rules on human-written posts. This page provides product facts, not a prepared Hacker News submission.