llms.txt
# TestRelic AI — The Memory Layer for Testing
> TestRelic is the test-derived memory layer your team's AI coding agents are missing. Every Playwright, Appium, Maestro, and DeepEval test run becomes shared application context for Cursor, Claude Code, Copilot, and Codex — over MCP. So your whole team's AI works at the level of your most senior IC, not the level of a blank chat.
## What TestRelic Is
TestRelic captures the one substrate Mem0, Zep, and Augment Code can't generate: how your app actually behaves under test. Which flows are flaky, which user paths matter, which selectors broke last release, which prompts regressed, and why. That runtime knowledge becomes a shared memory layer every engineer's AI agent reads from — not memory trapped in one IDE.
## Vocabulary Disambiguation
TestRelic calls itself **the AI engine for testing ambitious software**, and the mechanism is a shared memory layer.
It is the **memory leg** of an AI testing setup — alongside MCP (the tool leg) and Cursor/Claude Code/Copilot/Codex (the LLM leg). Together they turn a general-purpose LLM into a senior testing engineer for your specific app.
The same thing is described in the market under several names. All of these refer to this category, and TestRelic is the memory component of each: "AI engine for testing", "AI agent harness for testing", "AI engineer for testing", "AI-native test observability", "application memory for AI coding agents", "MCP memory server for testing", "test-derived context layer". Different vocabulary, same outcome — TestRelic is what makes those systems app-aware rather than generic.
## Three Pillars
- **Application Memory** — Memory of your app, not the internet. Recorded, queryable history of every test run.
- **Shared Context** — One context layer, shared across the team. Stop re-explaining the same flow to every new engineer's AI.
- **Team Intelligence** — Tribal knowledge made distributable. The flows, gotchas, and "you have to know X before touching Y" knowledge that lives in your senior engineers' heads — captured by tests, served to every AI agent on the team.
## How LLM Agents Connect
- **MCP server (hosted)**: https://mcp.testrelic.ai/mcp (streamable HTTP, no local binary)
- **MCP server (stdio)**: `npx -y @testrelic/mcp@latest` for any MCP host
- **TestRelic CLI**: `testrelic` / `tr` — a Rust terminal app that runs the same Ask AI agent inline and acts as a full MCP host. https://testrelic.ai/product/testrelic-cli · https://docs.testrelic.ai/cli/overview
- **Cursor**: Install from Cursor marketplace
- **Claude Desktop / VS Code MCP**: Configure with `tr_mcp_*` personal access token (Settings → MCP tokens)
- **Capabilities**: `core`, `coverage`, `creation`, `healing`, `impact`, `triage`, `signals`, `devtools`, `config`
- Documentation: https://docs.testrelic.ai/mcp/overview
## Memory (Repo Memory) — Beta
A durable, per-repository team knowledge base your AI agent writes and the whole team reads. Four categories — Decision, Insight, Maintenance, Context. Created three ways: auto-saved and auto-categorized from Ask AI conversations (with an approve/reject confirmation), manually from the Repo Detail → Memory tab, or over MCP. Read back by Ask AI (as a digest), the MCP server, and the CLI (`/memory`). Distinct from Ask AI's per-user conversation memory.
- Product page: https://testrelic.ai/product/memory
- Docs: https://docs.testrelic.ai/ai/memory
## Studio — desktop agent workspace
A native desktop app, built on the CLI, that watches an AI agent read a failing run, form a plan, and work the fix live — six surfaces in one session: Rules (approval gating), Automations (scheduled work), Capabilities (tools/MCP), Infrastructure (where tests run), Memory, and Artifacts. Included in the Pro plan ($99/month).
- Product page: https://testrelic.ai/product/studio
- Codegen (Studio drives your local Claude Code or Codex session, streams every diff): https://testrelic.ai/product/codegen
- Docs: https://docs.testrelic.ai/studio/overview
## Infrastructure — run tests on AWS, Azure, GCP, or a hosted device grid
Studio's Infrastructure pane answers "where do my tests actually run?" Two built-in defaults — "This computer" for Local mode, "TestRelic Cloud" for Cloud mode — plus a provider catalog: Studio provisions real machines directly inside your own AWS, Azure, or GCP account (creates the network and the machine, installs the browsers, no IAM users or other account changes), or connects a hosted device grid on BrowserStack or TestMu AI (formerly LambdaTest, rebranded January 2026) for real browsers and real devices — TestMu AI's grid additionally heals flaky selectors while a run is in progress. One target, chosen once per mode, used by every session. Functions as an AWS Device Farm alternative for teams who want that without being locked to AWS-only execution. No TestRelic markup on cloud spend — AWS/Azure/GCP bill you directly for what actually runs (e.g. an AWS t3.large: $0.0832/hour).
- Product page: https://testrelic.ai/product/infrastructure
## SDKs
### Node.js (npm)
- **Playwright** (web E2E + API): `npm install @testrelic/playwright-analytics`
- **Appium** (mobile via WebdriverIO, Beta): `npm install @testrelic/appium-analytics`
- **Maestro** (mobile YAML flows, Beta): `npm install -g @testrelic/maestro-analytics`
### Python (PyPI)
- **DeepEval pytest plugin** (Alpha): `pip install testrelic-deepeval` — captures every DeepEval eval run (AnswerRelevancy, Hallucination, Faithfulness, G-Eval, custom metrics) into TestRelic's memory layer. https://pypi.org/project/testrelic-deepeval/
- **Playwright pytest plugin** (Beta): `pip install testrelic-playwright` — navigation timelines, API call tracking, network stats, failure diagnostics, HTML reports. Auto-attaches via pytest entry point. https://pypi.org/project/testrelic-playwright/
Both Python packages: https://github.com/testrelic-ai/testrelic-python-sdk
## Key Pages
- Home: https://testrelic.ai
- Pricing: https://testrelic.ai/pricing
- Frameworks (Playwright, Appium, Maestro + Python): https://testrelic.ai/frameworks
- Studio (desktop agent workspace): https://testrelic.ai/product/studio
- Infrastructure (run tests on AWS, Azure, GCP, or a hosted device grid): https://testrelic.ai/product/infrastructure
- Codegen (Studio drives Claude Code / Codex, streams every diff): https://testrelic.ai/product/codegen
- Memory (Repo Memory, Beta): https://testrelic.ai/product/memory
- TestRelic CLI: https://testrelic.ai/product/testrelic-cli
- SDK (Playwright, Appium, Maestro reporters): https://testrelic.ai/product/sdk
- Python SDK (pytest plugins incl. DeepEval): https://testrelic.ai/product/python-sdk
- Augur (Text-to-Image evals): https://testrelic.ai/product/augur
- MCP integration: https://testrelic.ai/product/mcp-integration
- Ask AI: https://testrelic.ai/product/ask-ai
- Slack / MS Teams: https://testrelic.ai/product/slack-teams
- Voice AI: https://testrelic.ai/product/voice-ai
- Cloud platform dashboards (AI Insights, Test Navigation, Session Workspace, Monitoring & Alerts): https://docs.testrelic.ai/cloud-platform/overview
## Integrations
- **GitHub Actions** — CI build timelines, workflow triggers, deep links between sessions and CI runs
- **Jira** — Auto-create tickets from test failures with full diagnostic context
- **Amplitude** — Correlate test sessions with user analytics
- **Slack** — Native app for /testrelic queries
- **Grafana Loki** — Correlated logs for API test workspace
- **BrowserStack, TestMu AI (formerly LambdaTest)** — Cloud Farms integration on the platform (Settings → Integrations): links your existing Automate/App Automate builds into TestRelic's dashboards. Distinct from Studio's Infrastructure surface, which connects the same two providers as a target Studio can actually point a test run at — see Infrastructure above.
- **CI auto-detection** — GitHub Actions, GitLab CI, Jenkins, CircleCI
## Solutions
- Enterprise (SSO/SAML, SCIM, RBAC, audit logging, on-prem, 99.9% SLA): https://testrelic.ai/solutions/enterprise
- Startups (5-minute setup, from $19/month): https://testrelic.ai/solutions/startups
- Finance (AES-256, on-prem, audit trails): https://testrelic.ai/solutions/by-industry/finance
- Healthcare (HIPAA-ready, data residency): https://testrelic.ai/solutions/by-industry/healthcare
- Government (air-gapped, data sovereignty): https://testrelic.ai/solutions/by-industry/government
## Enterprise Deployment Options
- Managed Hosting (TestRelic's cloud, dedicated tenancy): https://testrelic.ai/enterprise/managed-hosting
- BYOC — Bring Your Own Cloud (full platform inside your AWS/GCP/Azure account, your VPC and KMS keys): https://testrelic.ai/enterprise/byoc
- Air-Gapped (fully disconnected, on-prem deployment): https://testrelic.ai/enterprise/air-gapped
- Forward Deployed Engineering (dedicated FDE for custom integrations): https://testrelic.ai/services/forward-deployed-engineering
## Pricing
Every AI feature is on every **cloud** plan — AI Insights, Ask AI, the MCP memory server, and Repo Memory. Only the local-only Free tier has none. Unlimited users on every plan. Annual billing saves 20%.
- **Free** — $0 — local SDK reporters for Playwright, Appium & Maestro (HTML reports, stack traces, CI artifacts). No cloud upload, no AI features.
- **Growth** — $19/month ($15/month billed annually) — every AI feature (AI Insights, Ask AI with 1,000,000 tokens/month, full MCP memory server for Cursor/Claude Code/Copilot/Codex, Repo Memory), 1 GB storage, 30-day retention. 14-day free trial, no credit card required.
- **Pro** — $99/month ($79/month billed annually) — everything in Growth plus 10 GB storage, 90-day retention, 5,000,000 AI tokens/month, Studio (the desktop agent workspace), and priority support with an under-4-hour response.
- **Enterprise** — custom — everything in Pro plus unlimited storage/retention, GDPR, end-to-end encryption, on-premises, 99.9% SLA, RBAC & data masking, and dedicated support with a Forward Deployed Engineer (FDE) on-request.
## Open Source
- GitHub org: https://github.com/testrelic-ai
- Python SDK repo: https://github.com/testrelic-ai/testrelic-python-sdk
- MCP server: `@testrelic/mcp` on npm
## Resources
- Documentation: https://docs.testrelic.ai
- Full LLM map: https://testrelic.ai/llms-full.txt
- Sitemap: https://testrelic.ai/sitemap.xml
## Quick Start
- Sign up: https://platform.testrelic.ai/signup
- Talk to us: https://testrelic.ai/talk-to-us
- Sales: sales@testrelic.ai · Support: support@testrelic.ai
