MCP tools reference
Curated list of TestRelic MCP tools by capability—core, coverage, creation, healing, impact, triage, signals, devtools, ai, marketplace, apps, artifacts, sessions, and memory—with links to the full upstream catalog.
All tools are exposed to MCP clients with the tr_ prefix. Enable the matching capability or the tool will not be registered.
tr_list_reposList repositories you can access.tr_recent_runsRecent test runs with filters (status, framework).tr_describe_repoRepo metadata and integration summary.tr_get_configResolved MCP configuration (capabilities, paths).tr_healthConnectivity and cache health — good first call.tr_integration_statusLive check for a specific integration type.tr_fetch_cachedRetrieve a large payload by cache_key after truncation.tr_user_journeysTop user journeys (usage-backed).tr_coverage_gapsRanked gaps between journeys and tests.tr_coverage_reportProgress toward coverage targets.tr_test_mapMap tests to journeys and code nodes.tr_plan_testMarkdown plan from a journey or goal.tr_generate_testRunnable test code from a plan.tr_dry_run_testTypecheck / list tests on generated output.tr_generate_assertionStable assertion snippet for a step.tr_list_templatesAvailable framework templates.tr_heal_runProposed patch for a failing run.tr_suggest_locatorPrefer stable locators over brittle selectors.tr_replay_failureArtefacts and steps to reproduce offline.tr_analyze_diffBlast radius and tests touched by a diff.tr_risk_scoreLightweight risk signal from journey usage.tr_select_testsMUST / SHOULD / OPTIONAL buckets for testing.tr_diagnose_runCompact failure summary for a run.tr_compare_runsDiff two runs for regressions or fixes.tr_flaky_auditRank flaky tests over a window.tr_search_failuresText search across failures.tr_ai_rcaRoot-cause narrative when available.tr_suggest_fixPlatform suggestion for a failing test.tr_create_jiraCreate or dedupe a Jira issue.tr_dismiss_flakyMark known-flaky with a reason.tr_user_impactCorrelate a run with user- and log-based impact.tr_production_signalQuery production logs (LogQL) in a window.tr_affected_sessionsSessions cohort for comms or rollback.tr_project_trendsPass rate, duration, flakiness trends.tr_active_alertsActive platform alerts.tr_index_repoIndex a local checkout for semantic search.tr_search_codeVector search over indexed code.tr_cache_statsCache layer counters for debugging.tr_ai_catalogBrowse the Ask-AI tool catalog.tr_ai_askRun a single-turn Ask-AI agent request.tr_ai_conversationsList and manage Ask-AI conversations.tr_ai_generate_artifactGenerate dashboards, reports, test plans, presentations, navigation paths.tr_marketplace_catalogList available testing integrations.tr_marketplace_connectValidate / connect or start an OAuth flow.tr_marketplace_disconnectDisconnect a connected integration.tr_marketplace_invokeUnified invoke against a connected integration.tr_apps_listList connected apps.tr_apps_actionsList the actions available for an app.tr_apps_connectConnect or disconnect an app via OAuth.tr_apps_executeUniversal execute of an action on a connected app.tr_artifacts_listList artifacts produced by the Ask-AI agent.tr_artifacts_getFetch a single artifact.tr_artifacts_exportExport an artifact (PNG, PDF).tr_artifacts_saveSave an artifact to a file.tr_sessions_searchSearch cloud test sessions.tr_sessions_videoRetrieve session video.tr_sessions_screenshotsRetrieve session screenshots.tr_sessions_logsRetrieve session logs.tr_sessions_workspaceRender a multi-pane session workspace.tr_list_repo_memoriesList saved memories for a repository.tr_get_repo_memoryFetch a specific repo memory.tr_save_repo_memorySave a new repo memory (auto-categorized).Search or filter by capability to find a tool below — the same names and one-line purposes as the tables further down this page.
Deprecated aliases are opt-in
The tr_* names are the current v2 namespace — 66 tools after a dedupe pass removed duplicate and superseded names (down from 86). The older testrelic_* names are off by default; enable them only while migrating with --legacy-aliases or TESTRELIC_MCP_LEGACY_ALIASES=1. Prefer the tr_* form in new prompts and configs.
For the full, machine-maintained table (every tool and description), see the @testrelic/mcp README on GitHub.
core
| Tool | Purpose |
|---|---|
tr_list_repos | List repositories you can access. |
tr_recent_runs | Recent test runs with filters. Filter by status (including cancelled) and by framework (Playwright, pytest, and others). |
tr_describe_repo | Repo metadata and integration summary. |
tr_get_config | Resolved MCP configuration (capabilities, paths). |
tr_health | Connectivity and cache health—good first call. |
tr_integration_status | Live check for a specific integration type. |
tr_fetch_cached | Retrieve a large payload by cache_key when a result was truncated to the token budget. |
coverage
| Tool | Purpose |
|---|---|
tr_user_journeys | Top user journeys (usage-backed). |
tr_coverage_gaps | Ranked gaps between journeys and tests. |
tr_coverage_report | Progress toward coverage targets. |
tr_test_map | Map tests to journeys and code nodes. |
creation
| Tool | Purpose |
|---|---|
tr_plan_test | Markdown plan from a journey or goal. |
tr_generate_test | Runnable test code from a plan. |
tr_dry_run_test | Typecheck / list tests on generated output. |
tr_generate_assertion | Stable assertion snippet for a step. |
tr_list_templates | Available framework templates. |
healing
| Tool | Purpose |
|---|---|
tr_heal_run | Proposed patch for a failing run. |
tr_suggest_locator | Prefer stable locators over brittle selectors. |
tr_replay_failure | Artefacts and steps to reproduce offline. |
impact
| Tool | Purpose |
|---|---|
tr_analyze_diff | Blast radius and tests touched by a diff. |
tr_risk_score | Lightweight risk signal from journey usage. |
tr_select_tests | MUST / SHOULD / OPTIONAL buckets for testing. |
triage
| Tool | Purpose |
|---|---|
tr_diagnose_run | Compact failure summary for a run. Run status reflects the run outcome, so a run containing failures is never reported as passed. |
tr_compare_runs | Diff two runs for regressions or fixes. |
tr_flaky_audit | Rank flaky tests over a window. |
tr_search_failures | Text search across failures. |
tr_ai_rca | Root-cause narrative when available. |
tr_suggest_fix | Platform suggestion for a failing test. |
tr_create_jira | Create or dedupe a Jira issue. |
tr_dismiss_flaky | Mark known-flaky with a reason. |
signals
| Tool | Purpose |
|---|---|
tr_user_impact | Correlate a run with user- and log-based impact. |
tr_production_signal | Query production logs (LogQL) in a window. |
tr_affected_sessions | Sessions cohort for comms or rollback. |
devtools
| Tool | Purpose |
|---|---|
tr_project_trends | Pass rate, duration, flakiness trends. Accepts a days window and returns the total run count for the window. |
tr_active_alerts | Active platform alerts. |
tr_index_repo | Index a local checkout for semantic search. |
tr_search_code | Vector search over indexed code. |
tr_cache_stats | Cache layer counters for debugging. |
ai
Enable with --caps ai. Exposes the Ask-AI tool catalog, a single-turn agent, conversation management, and artifact generators.
| Tool | Purpose |
|---|---|
tr_ai_catalog | Browse the Ask-AI tool catalog. |
tr_ai_ask | Run a single-turn Ask-AI agent request. |
tr_ai_conversations | List and manage Ask-AI conversations. |
tr_ai_generate_artifact | Generate dashboards, reports, test plans, presentations, or navigation paths. |
marketplace
Enable with --caps marketplace. First-class testing integrations (Jira, GitHub Actions, BrowserStack, LambdaTest, Sentry, Amplitude, Grafana Loki).
| Tool | Purpose |
|---|---|
tr_marketplace_catalog | List available testing integrations. |
tr_marketplace_connect | Validate / connect or start an OAuth flow for an integration. |
tr_marketplace_disconnect | Disconnect a connected integration. |
tr_marketplace_invoke | Unified invoke against a connected integration. |
apps
Enable with --caps apps. Generic action runner for connected apps (Slack, Notion, Linear, HubSpot, Google Calendar, …).
| Tool | Purpose |
|---|---|
tr_apps_list | List connected apps. |
tr_apps_actions | List the actions available for an app. |
tr_apps_connect | Connect or disconnect an app via OAuth. |
tr_apps_execute | Universal execute of an action on a connected app. |
artifacts
Enable with --caps artifacts. Manage artifacts produced by the Ask-AI agent.
| Tool | Purpose |
|---|---|
tr_artifacts_list | List artifacts produced by the Ask-AI agent. |
tr_artifacts_get | Fetch a single artifact. |
tr_artifacts_export | Export an artifact (PNG, PDF). |
tr_artifacts_save | Save an artifact to a file. |
sessions
Enable with --caps sessions. Cloud test session workspaces (BrowserStack / LambdaTest).
| Tool | Purpose |
|---|---|
tr_sessions_search | Search cloud test sessions. |
tr_sessions_video | Retrieve session video. |
tr_sessions_screenshots | Retrieve session screenshots. |
tr_sessions_logs | Retrieve session logs. |
tr_sessions_workspace | Render a multi-pane session workspace. |
memory
Enable with --caps memory. Read and write repo memory — durable, repo-scoped context (project map, conventions, prior decisions) that gives any assistant precise context on the next turn.
| Tool | Purpose |
|---|---|
tr_list_repo_memories | List saved memories for a repository. |
tr_get_repo_memory | Fetch a specific repo memory. |
tr_save_repo_memory | Save a new repo memory (auto-categorized). |
Security
Disclosure and versioning policy for the upstream project: SECURITY.md.
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.
Monitoring Overview
An overview of TestRelic's monitoring dashboards — Smoke, Regression, and Nightly Observatory — for tracking automated quality gates and test health trends.