MCP capabilities
Enable TestRelic MCP tool groups with --caps—core, coverage, creation, healing, impact, triage, signals, devtools, ai, marketplace, apps, artifacts, sessions, and memory—and pick a bundle that matches your workflow.
Tools are grouped into capabilities. The core group is always available; others are enabled by passing a comma-separated list to --caps (or TESTRELIC_MCP_CAPS).
Click a capability to toggle it, or pick a suggested bundle — the --caps string and tool count update to match, the same way starting the server with a different --caps list would.
| Capability | What you get |
|---|---|
| core | Projects, recent runs, resolved config, health, integration status. |
| coverage | User journeys, test map, coverage gaps, coverage report, cached payloads. |
| creation | Plan → generate tests, templates, dry-run, assertion helpers. |
| healing | Patch proposals, locator suggestions, replay plans for failures. |
| impact | Diff analysis, risk score, MUST/SHOULD/OPTIONAL test selection. |
| triage | Diagnose runs, flaky audit, compare runs, RCA, Jira create/dedupe, dismiss flaky, search failures. |
| signals | User impact, production logs (Loki), affected sessions. |
| devtools | Project trends, alerts, semantic code search, repo indexing, cache stats. |
| ai | Ask-AI tool catalog, single-turn agent, conversation management, and artifact generators (dashboards, reports, test plans, presentations, navigation paths). |
| marketplace | First-class testing integrations (Jira, GitHub Actions, BrowserStack, LambdaTest, Sentry, Amplitude, Grafana Loki): catalog, validate/connect/OAuth-start, disconnect, and a unified invoke. |
| apps | Generic action runner for connected apps (Slack, Notion, Linear, HubSpot, Google Calendar, …): list apps, list actions, connect/disconnect via OAuth, and a universal execute. |
| artifacts | List, get, export (PNG, PDF), and save-to-file for artifacts produced by the Ask-AI agent. |
| sessions | Cloud test session workspaces (BrowserStack / LambdaTest): search, video, screenshots, logs, and a multi-pane workspace render. |
| memory | Repo memory: read, list, and save durable, repo-scoped context (project map, conventions, prior decisions) so any assistant gets precise context on the next turn. |
Suggested bundles
- Day-to-day authoring:
core,coverage,creation,healing,impact - On-call / failures: add
triage,signals - Repo-wide exploration: add
devtools - Ask-AI & artifacts: add
ai,artifactsfor the Ask-AI agent plus generating and exporting dashboards, reports, and test plans - Integrations & connected apps: add
marketplace,appsto wire testing integrations and run actions on connected apps - Cloud sessions: add
sessionsto open BrowserStack / LambdaTest session workspaces - Repo memory: add
memoryto read and save durable repo-scoped context across sessions
Example:
--caps core,coverage,creation,healing,impact,triageEnable only what you need so your assistant sees a smaller tool schema.
Next steps
MCP authentication
Create a TestRelic MCP personal access token, store it with the login CLI or TESTRELIC_MCP_TOKEN, and align TESTRELIC_CLOUD_URL with your cloud API base when needed.
MCP and your testing suite
How TestRelic SDK reporters, the cloud platform, and the MCP server fit together—runs upload from CI and local, while the MCP gives your IDE assistant structured access to that data.