{ "data": { "a": { "slug": "amplitude", "name": "Amplitude", "vendor": "Amplitude, Inc.", "vendorUrl": "https://amplitude.com", "kind": "http-api", "category": "product-analytics", "summary": "Amplitude is a hosted product analytics platform that records user events and answers questions about funnels, retention, cohorts, experiments and feature flags. Agents reach it through an official remote MCP server, REST APIs and the `amp` command line.", "url": "https://www.anchorterminal.com/tools/amplitude", "markdownUrl": "https://www.anchorterminal.com/tools/amplitude.md", "slimMarkdownUrl": "https://www.anchorterminal.com/tools/amplitude.min.md", "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/amplitude.json", "dns": "amplitude.grade.anchorterminal.com", "repo": "https://github.com/amplitude/mcp-marketplace", "license": "Proprietary service under Amplitude's terms of service. The SDKs, the `amp` CLI and the MCP marketplace plugins on GitHub and npm are MIT", "transports": [ "http", "streamable-http" ], "remoteUrl": "https://mcp.amplitude.com/mcp", "packages": [ { "registry": "npm", "name": "@amplitude/developer-cli" }, { "registry": "npm", "name": "@amplitude/analytics-node" }, { "registry": "pypi", "name": "amplitude-analytics" } ], "auth": "mixed", "authNotes": "Access is self-serve, with no app review or partner approval. The MCP server takes only OAuth 2.0 (authorisation code with PKCE S256, dynamic client registration, scopes `mcp:read`, `mcp:write` and `offline_access`), and a person signs in to Amplitude in a browser. Each call runs with that user's permissions, limited by two role actions, Use MCP (read) and Use MCP (write), per project. The Developer API takes a Bearer token from the OAuth device flow or a personal access token with eight scopes. The older REST APIs use HTTP Basic with a project API key and secret key, the Experiment Management API a management key, and HTTP V2 ingestion the public project API key in the request body.", "pricing": "freemium", "pricingNotes": "Free plan with 2 million events a month and no card, and the pricing page says MCP and agent access work on every plan. Plus is usage-based from $0 with the first 2 million events free, up to 70 million, and needs a card to see an estimate. Growth and Enterprise are priced by sales. API and MCP calls aren't metered in money. There's no separate sandbox, so an agent starts on a Free organisation (https://amplitude.com/pricing, checked 2026-10-07).", "priceSummary": "Freemium", "where": "hosted", "x402": { "level": "no", "evidence": "No x402, MPP or L402 in the API docs, the Developer API's OpenAPI document or the pricing page (checked 2026-10-07).", "endpoints": [] }, "toolCount": 45, "popularity": { "githubStars": null, "npmWeekly": 1331780, "pypiWeekly": 1779665, "asOf": "2026-10-07" }, "docsUrl": "https://amplitude.com/docs/apis", "llmsTxt": "https://amplitude.com/docs/llms.txt", "openapi": "https://developer-api.amplitude.com/openapi.yaml", "registryName": "com.amplitude/mcp-server", "capabilities": [ "analytics.query", "analytics.funnels", "analytics.experiments", "analytics.flags", "analytics.events" ], "tags": [ "official", "hosted", "mcp", "oauth", "openapi", "llms-txt", "freemium", "free-tier", "no-card", "eu-region", "cli", "status-page", "soc2", "python", "typescript" ], "lastRelease": "2026-10-29", "graded": true, "anchor": { "graded": true, "score": 69.3, "grade": "B", "agentReady": false, "rank": 249, "ranked": true, "rankOf": 1486, "categoryRank": 3, "methodology": "0.7", "run": "2026-10-01", "scores": { "ergonomics": 72, "maintenance": 83, "payments": 30, "reliability": 75, "schema": 72, "security": 74, "transparency": 79 }, "pending": [ "performance", "tasks" ], "assessment": { "confidence": "medium", "date": "2026-11-01" }, "negative": 0, "verdict": "The remote MCP server covers queries, funnels, retention, experiments and flags under OAuth with PKCE, read and write scopes and per-project role actions, and the Free plan needs no card. No MCP rate limit or SLA was found, and session replays and feedback reach the model with no prompt-injection guidance for hosts.", "bestFor": "A team already on Amplitude that wants an agent to query charts, funnels, retention and experiment results, edit dashboards and manage the tracking plan under its own permissions.", "strengths": [ "Remote MCP server with 44 documented tools for queries, funnels, retention, cohorts, experiments, flags and the tracking plan, on US and EU hosts", "OAuth with PKCE, dynamic client registration, a revocation endpoint and `mcp:read` and `mcp:write` scopes, plus two role actions enforced per project on every call", "Progressive discovery (`?discovery=progressive`) starts with a small tool list and loads schemas on demand", "Deletes need a second call with `confirmed` set to true, and cohort syncs preview the export before sending", "Free plan with 2 million events a month and no card, and the pricing page says MCP works on every plan" ], "weaknesses": [ "No rate limit for the MCP server or the Developer API was found in the reviewed documentation", "status.amplitude.com lists 23 incidents between 3 August and 1 November 2026, mostly ingestion and export delays, and a 13 July outage of the UI, MCP and REST APIs", "No SLA found, and the terms of 21 July 2026 disclaim uninterrupted service", "MCP is on by default for every user, and the Member, Manager and Admin roles include write access until an admin changes them", "Session replays, feedback comments and user properties reach the model with no prompt-injection guidance for hosts" ], "agentNotes": [ "Connect to https://mcp.amplitude.com/mcp, or https://mcp.eu.amplitude.com/mcp for EU projects. The client must support streaming HTTP and OAuth, since API keys aren't accepted", "Append `?discovery=progressive` to load tool schemas on demand, then call `get_amplitude_context` first to pick the project and read its AI context", "Ask an admin for a role with only Use MCP (read) on the projects you need. Tools stay listed and fail at call time when the role lacks the action", "Send events through the HTTP V2 API with an `insert_id` on each one. On 429, pause that user or device for 30 seconds before retrying", "Budget Dashboard REST queries by cost (days times conditions times chart cost), within 108,000 an hour and 5 concurrent requests per project" ], "metrics": { "kind": "remote", "measured": false }, "reviewCount": 4, "avgRating": 3, "history": [ { "basis": "public evidence", "confidence": "medium", "grade": "B", "methodology": "0.7", "pending": [ "performance", "tasks" ], "run": "2026-10-01", "runLabel": "October 2026 research run", "score": 69.3 } ], "editorialScores": { "ergonomics": 72, "maintenance": 83, "payments": 30, "reliability": 75, "schema": 72, "security": 74, "transparency": 63 }, "provenanceScore": 94, "checked": { "categories": { "ergonomics": "2026-11-01", "maintenance": "2026-11-01", "payments": "2026-11-01", "reliability": "2026-11-01", "schema": "2026-11-01", "security": "2026-11-01", "transparency": "2026-11-01" }, "last": "2026-11-01", "lastMoved": "2026-11-01", "oldest": "2026-11-01", "stale": false, "staleAfterDays": 90 } }, "connect": { "install": "npm install -g @amplitude/developer-cli", "http": "curl --location --request GET 'https://amplitude.com/api/3/chart/:chart_id/csv' \\\n-u '{api_key}:{secret_key}'", "claudeCode": "claude mcp add -t http -s user Amplitude \"https://mcp.amplitude.com/mcp\"", "config": { "mcpServers": { "amplitude": { "type": "http", "url": "https://mcp.amplitude.com/mcp" } } } }, "letme": { "capability": "https://letme.dev/analytics.query", "tool": "https://letme.dev/amplitude" }, "area": "business", "provenance": { "legalEntity": "Amplitude, Inc.", "domain": "amplitude.com", "domainRegistered": "1996-05-09", "endpointOnVendorDomain": true, "terms": "https://amplitude.com/terms", "privacy": "https://amplitude.com/privacy", "statusPage": "https://status.amplitude.com", "changelog": "https://amplitude.com/releases", "securityTxt": "valid", "checked": "2026-10-07", "notes": [ "The terms of service (effective 21 July 2026) and the privacy notice (updated 31 August 2026) name Amplitude, Inc., 201 3rd Street, Suite 200, San Francisco, CA 94103.", "The MCP server answers at mcp.amplitude.com and mcp.eu.amplitude.com, the Developer API at developer-api.amplitude.com and ingestion at api2.amplitude.com, all amplitude.com subdomains.", "amplitude.com/.well-known/security.txt gives a Bugcrowd tracker and security@amplitude.com as contacts and expires 2028-12-31.", "RDAP for amplitude.com gives a registration date of 1996-05-09.", "The data processing addendum for the terms of service is dated 4 June 2024 and points to the sub-processor list at amplitude.com/subprocessor-list, updated 13 May 2026." ], "score": 94 }, "pageJsonUrl": "https://www.anchorterminal.com/tools/amplitude.json", "live": { "slug": "amplitude", "probe": { "target": "https://mcp.amplitude.com/mcp", "method": "get", "lastAt": "2026-11-07T18:20:10.354607203Z", "lastOk": true, "lastStatus": 401, "lastMs": 695, "lastNote": "asks for credentials", "authRequired": true, "uptime24h": 100, "uptime30d": 99.99, "p50ms24h": 602, "p95ms24h": 931, "samples24h": 239, "samples30d": 7463, "days": [ { "date": "2026-10-08", "probes": 93, "ok": 93 }, { "date": "2026-10-09", "probes": 250, "ok": 250 }, { "date": "2026-10-10", "probes": 247, "ok": 247 }, { "date": "2026-10-11", "probes": 247, "ok": 247 }, { "date": "2026-10-12", "probes": 247, "ok": 247 }, { "date": "2026-10-13", "probes": 248, "ok": 248 }, { "date": "2026-10-14", "probes": 249, "ok": 249 }, { "date": "2026-10-15", "probes": 253, "ok": 253 }, { "date": "2026-10-16", "probes": 252, "ok": 252 }, { "date": "2026-10-17", "probes": 253, "ok": 253 }, { "date": "2026-10-18", "probes": 256, "ok": 256 }, { "date": "2026-10-19", "probes": 255, "ok": 254 }, { "date": "2026-10-20", "probes": 252, "ok": 252 }, { "date": "2026-10-21", "probes": 253, "ok": 253 }, { "date": "2026-10-22", "probes": 253, "ok": 253 }, { "date": "2026-10-23", "probes": 248, "ok": 248 }, { "date": "2026-10-24", "probes": 251, "ok": 251 }, { "date": "2026-10-25", "probes": 250, "ok": 250 }, { "date": "2026-10-26", "probes": 252, "ok": 252 }, { "date": "2026-10-27", "probes": 249, "ok": 249 }, { "date": "2026-10-28", "probes": 251, "ok": 251 }, { "date": "2026-10-29", "probes": 249, "ok": 249 }, { "date": "2026-10-30", "probes": 248, "ok": 248 }, { "date": "2026-10-31", "probes": 251, "ok": 251 }, { "date": "2026-11-01", "probes": 249, "ok": 249 }, { "date": "2026-11-02", "probes": 250, "ok": 250 }, { "date": "2026-11-03", "probes": 248, "ok": 248 }, { "date": "2026-11-04", "probes": 246, "ok": 246 }, { "date": "2026-11-05", "probes": 185, "ok": 185 }, { "date": "2026-11-06", "probes": 241, "ok": 241 }, { "date": "2026-11-07", "probes": 187, "ok": 187 } ] }, "vendorStatus": { "page": "https://status.amplitude.com", "indicator": "unknown", "summary": "no machine-readable status found", "checkedAt": "2026-11-07T04:47:06.124166931Z" }, "versions": [ { "registry": "mcp-registry", "name": "com.amplitude/mcp-server", "version": "1.0.0", "seenAt": "2026-11-07T18:16:28.95801333Z" }, { "registry": "npm", "name": "@amplitude/analytics-node", "version": "1.5.83", "seenAt": "2026-11-06T23:06:09.538599909Z" }, { "registry": "npm", "name": "@amplitude/developer-cli", "version": "0.7.0", "seenAt": "2026-11-06T23:06:07.683421691Z" }, { "registry": "pypi", "name": "amplitude-analytics", "version": "1.3.0", "released": "2026-10-15", "seenAt": "2026-11-06T23:06:11.527314522Z" } ], "githubStars": 46, "npmWeekly": 1844, "pypiWeekly": 2063132, "securityTxt": { "url": "https://amplitude.com/.well-known/security.txt", "state": "valid", "expires": "2028-12-31T23:59:00Z", "checkedAt": "2026-11-07T15:59:54.27374171Z" }, "llmsTxt": { "url": "https://amplitude.com/docs/llms.txt", "ok": true, "status": 200, "checkedAt": "2026-11-07T15:31:33.274986966Z" }, "domain": { "domain": "amplitude.com", "registered": "1996-05-09", "source": "https://rdap.verisign.com/com/v1/domain/amplitude.com", "checkedAt": "2026-11-01T14:23:33.372388831Z" }, "pages": [ { "url": "https://amplitude.com/releases", "kind": "changelog", "status": 200, "checkedAt": "2026-11-07T13:59:12.996169879Z", "changedAt": "2026-11-04T08:14:55.402246886Z", "fingerprint": "1d5def6f5d23" }, { "url": "https://amplitude.com/docs/api/content/faq/snowflake-password-auth-deprecation", "kind": "deprecations", "status": 200, "checkedAt": "2026-11-07T13:59:06.467486166Z", "changedAt": "0001-01-01T00:00:00Z", "fingerprint": "ff68ac9dbf41" }, { "url": "https://amplitude.com/pricing", "kind": "pricing", "status": 200, "checkedAt": "2026-11-07T13:59:08.799284299Z", "changedAt": "2026-11-07T13:59:08.799284299Z", "fingerprint": "fe54b37d50e6" }, { "url": "https://amplitude.com/privacy", "kind": "privacy", "status": 200, "checkedAt": "2026-11-07T13:59:11.280729948Z", "changedAt": "2026-11-04T08:14:53.644961002Z", "fingerprint": "43b826abc7f7" }, { "url": "https://amplitude.com/terms", "kind": "terms", "status": 200, "checkedAt": "2026-11-07T13:59:15.278756091Z", "changedAt": "2026-10-30T05:31:27.547084835Z", "fingerprint": "41070f67d073" } ], "mcpTools": { "url": "https://mcp.amplitude.com/mcp", "checkedAt": "2026-11-07T01:42:21.529548873Z", "status": "auth", "note": "asks for credentials before listing its tools", "changedAt": "2026-11-07T01:42:21.529548873Z" }, "github": { "repo": "amplitude/mcp-marketplace", "checkedAt": "2026-11-02T08:30:21.0199871Z", "stars": 46, "archived": false, "issuesEnabled": true, "pushedAt": "2026-10-31T01:51:45Z", "openIssues": 4, "openPRs": 9, "issuesOpened90d": 3, "issuesClosed90d": 1, "prsMerged90d": 17, "ci": { "branch": "main", "runs": 31, "success": 31, "failure": 0, "other": 0, "latest": "success", "latestAt": "2026-10-30T13:06:02Z" }, "replies": { "sampled": 5, "unanswered": 3, "medianDays": 7.6 } }, "updatedAt": "2026-11-07T18:20:10.354607203Z" } }, "answer": "Amplitude and Harness Feature Management \u0026 Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management \u0026 Experimentation leads on reliability, agent ergonomics and security \u0026 auth.", "b": { "slug": "harness-fme", "name": "Harness Feature Management \u0026 Experimentation", "vendor": "Harness Inc.", "vendorUrl": "https://www.harness.io", "kind": "http-api", "category": "product-analytics", "summary": "Harness Feature Management \u0026 Experimentation (FME), formerly Split, manages feature flags, targeting, metrics and experiments. Agents reach it through the Admin API at api.split.io and the FME toolset of the Harness MCP server.", "url": "https://www.anchorterminal.com/tools/harness-fme", "markdownUrl": "https://www.anchorterminal.com/tools/harness-fme.md", "slimMarkdownUrl": "https://www.anchorterminal.com/tools/harness-fme.min.md", "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/harness-fme.json", "dns": "harness-fme.grade.anchorterminal.com", "repo": "https://github.com/harness/mcp-server", "transports": [ "http", "stdio", "streamable-http" ], "remoteUrl": "https://mcp.harness.io/mcp", "packages": [ { "registry": "npm", "name": "harness-mcp-v2" } ], "auth": "mixed", "authNotes": "The Admin API takes an Admin API key as a Bearer token, or a Harness personal access token (PAT) or service account token (SAT) in `x-api-key`. Admin keys carry fixed roles that can be combined and cannot be edited after creation, and a key cannot create another with more roles. The local MCP server reads HARNESS_API_KEY. The hosted MCP endpoint uses Harness Platform OAuth and must be enabled per account by Harness Support.", "pricing": "freemium", "pricingNotes": "An FME Free Plan is enabled from inside a Harness account; its limits are not published. The Enterprise plan and FME pricing are by quotation through sales. No FME prices appear on harness.io/pricing.", "priceSummary": "Freemium", "where": "both", "x402": { "level": "no", "evidence": "No x402, MPP or L402 in the FME docs, the Split API reference or harness.io/pricing (checked 2026-11-06).", "endpoints": [] }, "toolCount": 11, "popularity": { "githubStars": null, "npmWeekly": 8800, "pypiWeekly": null, "asOf": "2026-11-06" }, "docsUrl": "https://developer.harness.io/feature-management-experimentation/feature-management-and-experimentation-documentation", "llmsTxt": "https://developer.harness.io/llms.txt", "capabilities": [ "analytics.flags", "analytics.experiments", "analytics.events" ], "tags": [ "official", "hosted", "mcp", "remote-mcp", "oauth", "llms-txt", "free-tier", "sales-led", "status-page", "soc2", "iso27001", "typescript", "python", "cli", "closed-source" ], "lastRelease": "2026-11-04", "graded": true, "anchor": { "graded": true, "score": 69.2, "grade": "B", "agentReady": false, "rank": 257, "ranked": true, "rankOf": 1486, "categoryRank": 4, "methodology": "0.7", "run": "2026-10-01", "scores": { "ergonomics": 83, "maintenance": 81, "payments": 15, "reliability": 80, "schema": 71, "security": 79, "transparency": 61 }, "pending": [ "performance", "tasks" ], "assessment": { "confidence": "medium", "date": "2026-11-06" }, "negative": 0, "verdict": "Harness FME, formerly Split, gives agents 11 consolidated MCP tools with an FME-only toolset, a read-only switch and confirmation on writes, over an Admin API with documented limits and role-scoped keys. FME prices are not published, the free plan's limits are unstated, and the hosted MCP endpoint must be enabled by Harness Support.", "bestFor": "Teams already on Harness, or Split customers, who want agents to inspect and change flags and segments with guard rails.", "strengths": [ "11 generic MCP tools with `--toolsets=fme`, readOnlyHint on reads and elicitation before writes", "HARNESS_READ_ONLY blocks every create, update, delete and execute call", "Admin API keys with fixed, combinable roles, plus Harness PATs and SATs", "Documented rate limit with remaining and reset headers, and automatic backoff in the MCP server", "SOC 2 Type II, ISO 27001 and FedRAMP 20x, with FME audit logs and MCP audit sinks" ], "weaknesses": [ "No FME prices are published; Enterprise is by quotation and free plan limits are unstated", "The hosted MCP endpoint must be enabled per account by Harness Support", "The FME MCP docs page still describes building a Go binary, unlike the current npm package", "About 20 Admin API requests per 10 seconds per organisation, which large lists can hit", "No public OpenAPI file for the Admin API was found" ], "agentNotes": [ "Start the MCP server with `--toolsets=fme` so only FME resource types load", "Call `harness_describe(resource_type='fme_feature_flag')` before a create or update to see the fields", "Filter fme_feature_flag lists by name, tags or rollout status rather than listing every flag", "Set HARNESS_READ_ONLY for analysis tasks; leave confirmation on for kill, restore and delete", "On 429, wait at least the larger X-RateLimit-Reset-Seconds value before retrying" ], "metrics": { "kind": "remote", "measured": false }, "reviewCount": 0, "avgRating": 0, "history": [ { "basis": "public evidence", "confidence": "medium", "grade": "B", "methodology": "0.7", "pending": [ "performance", "tasks" ], "run": "2026-10-01", "runLabel": "October 2026 research run", "score": 69.2 } ], "editorialScores": { "ergonomics": 83, "maintenance": 81, "payments": 15, "reliability": 80, "schema": 71, "security": 79, "transparency": 60 }, "provenanceScore": 61, "checked": { "categories": { "ergonomics": "2026-11-06", "maintenance": "2026-11-06", "payments": "2026-11-06", "reliability": "2026-11-06", "schema": "2026-11-06", "security": "2026-11-06", "transparency": "2026-11-06" }, "last": "2026-11-06", "oldest": "2026-11-06", "stale": false, "staleAfterDays": 90 } }, "connect": { "install": "npx harness-mcp-v2@latest", "config": { "mcpServers": { "harness": { "args": [ "harness-mcp-v2@latest", "--toolsets=fme" ], "command": "npx", "env": { "HARNESS_API_KEY": "pat.xxx" } } } } }, "letme": { "capability": "https://letme.dev/analytics.flags", "tool": "https://letme.dev/harness-fme" }, "area": "business", "provenance": { "legalEntity": "Harness Inc.", "domain": "harness.io", "domainRegistered": "", "endpointOnVendorDomain": false, "terms": "https://www.harness.io/legal/subscription-terms", "privacy": "https://www.harness.io/legal/privacy", "statusPage": "https://status.harness.io", "changelog": "https://github.com/harness/mcp-server/releases", "securityTxt": "none", "checked": "2026-11-06", "notes": [ "The Admin API is on api.split.io, the former Split domain, and the hosted MCP endpoint on mcp.harness.io.", "The privacy statement names Harness Inc. as controller and Harness France SAS as EU representative. Its header says last updated 23 October 2026 and its body says version of 14 October 2026.", "developer.harness.io/robots.txt allows all and carries Content-Signal ai-train=yes, search=yes, ai-input=yes. docs.split.io allows all but edit and login paths.", "www.harness.io/.well-known/security.txt answered with an invalid-request page. The security page gives security@harness.io and a private bug bounty on request.", "RDAP gave no registration date for harness.io." ], "score": 61 }, "pageJsonUrl": "https://www.anchorterminal.com/tools/harness-fme.json", "live": { "slug": "harness-fme", "probe": { "target": "https://mcp.harness.io/mcp", "method": "get", "lastAt": "2026-11-07T18:20:29.454967067Z", "lastOk": true, "lastStatus": 401, "lastMs": 211, "lastNote": "asks for credentials", "authRequired": true, "uptime24h": 100, "uptime30d": 100, "p50ms24h": 195, "p95ms24h": 280, "samples24h": 200, "samples30d": 200, "days": [ { "date": "2026-11-06", "probes": 13, "ok": 13 }, { "date": "2026-11-07", "probes": 187, "ok": 187 } ] }, "vendorStatus": { "page": "https://status.harness.io", "indicator": "none", "summary": "All Systems Operational", "checkedAt": "2026-11-07T18:21:14.597545686Z" }, "versions": [ { "registry": "github", "name": "harness/mcp-server", "version": "v3.2.40", "released": "2026-11-05", "seenAt": "2026-11-06T23:32:13.419387707Z" }, { "registry": "npm", "name": "harness-mcp-v2", "version": "3.2.40", "seenAt": "2026-11-06T23:32:12.430886041Z" } ], "githubStars": 112, "npmWeekly": 8800, "securityTxt": { "url": "https://harness.io/.well-known/security.txt", "state": "none", "checkedAt": "2026-11-07T16:01:19.957252225Z" }, "llmsTxt": { "url": "https://developer.harness.io/llms.txt", "ok": true, "status": 200, "checkedAt": "2026-11-07T15:32:41.749004752Z" }, "pages": [ { "url": "https://www.harness.io/legal/privacy", "kind": "privacy", "status": 200, "checkedAt": "2026-11-07T14:29:30.551687912Z", "changedAt": "0001-01-01T00:00:00Z", "fingerprint": "b15935b42065" }, { "url": "https://www.harness.io/legal/subscription-terms", "kind": "terms", "status": 200, "checkedAt": "2026-11-07T14:29:35.70684084Z", "changedAt": "0001-01-01T00:00:00Z", "fingerprint": "871b29e12dd4" } ], "updatedAt": "2026-11-07T18:21:14.597545686Z" } }, "edges": [ { "by": 18, "category": "transparency", "leads": "amplitude", "name": "Transparency \u0026 trust", "reasons": { "a": "Closed service with terms effective 21 July 2026. The SDKs, CLI and marketplace plugins are MIT (15). The DPA of 4 June 2024 promises destruction within 90 days of termination and use only to run the service, while the terms let Amplitude apply machine learning to customer data to improve the services and delete data at its discretion after termination. The AI docs say data isn't used for model training (20 of 30). A written SDK lifecycle with at least 12 months of maintenance for a superseded major version. No deprecation policy for the REST APIs or MCP tools found (10 of 20). Sub-processor list updated 13 May 2026 with US and EU data centre locations and 10 business days' notice of additions under the DPA (18 of 20).", "b": "A closed service under Harness's subscription terms; the MCP server is MIT (18 of 30). The privacy statement covers website and sales data and defers customer data to the customer agreement and a published DPA; retention periods for customer data weren't found (18 of 30). A versioning policy for SDKs with support windows, and Split-era API endpoints marked deprecated with migration guides (14 of 20). Sub-processors are on the trust centre, which we didn't read, and the security page doesn't name data locations (10 of 20). 18+18+14+10 = 60." }, "scores": { "a": 79, "b": 61 } }, { "by": 15, "category": "payments", "leads": "amplitude", "name": "Payments \u0026 pricing", "reasons": { "a": "Read with the hosted rubric. No x402, MPP or L402 (0). Plans are public but Plus shows a price only through an estimator that asks for a card, and Growth and Enterprise go through sales (10). The Free plan has 2 million events a month with no card, and MCP works on it (20). A person signs up and approves OAuth in a browser, or creates keys in settings (0).", "b": "No x402, MPP or L402 (0). No FME prices are published; harness.io/pricing lists FME only under Enterprise, priced by quotation (0). The docs describe a Start FME Free Plan step inside a Harness account; its limits and whether a card is needed aren't stated (15 of 20). A person signs up for Harness in a browser and creates a key or token (0). 0+0+15+0 = 15." }, "scores": { "a": 30, "b": 15 } }, { "by": 11, "category": "ergonomics", "leads": "harness-fme", "name": "Agent ergonomics", "reasons": { "a": "44 tools is over the 30 line (5), with 10 added back for progressive discovery, which starts with four tools and loads schemas on demand, and a subset of about 25 for ChatGPT (15 of 25). Cursor and `limit` pagination on the Developer API (default 50, maximum 200), the Experiment Management API and the data warehouse tools, and filters on search and replay tools. The MCP page warns that hosts may truncate large charts (17 of 20). RFC 9457 errors with `error_code`, `retryable` and field errors on the Developer API and detailed 400 and 429 bodies on HTTP V2. MCP error shapes unread (15 of 20). `Idempotency-Key` and `dry_run` on some Developer API operations, `insert_id` on events, two-step deletes and optimistic concurrency on dashboards and notebooks. We couldn't see whether tools carry readOnlyHint or destructiveHint (13 of 20). Official SDKs in many languages cover ingestion and experiment evaluation, not queries, and the `amp` CLI prints compact JSON when piped (12 of 15).", "b": "11 consolidated tools, and `--toolsets=fme` narrows the server to FME (25). Flag lists filter server-side by name, tag and rollout status, and the Admin API paginates with markerLimit (17 of 20). Documented status codes, 429 headers, and a harness_diagnose tool. Error bodies aren't documented per endpoint (14 of 20). harness_list and harness_get carry readOnlyHint, write tools ask for confirmation by elicitation, and harness_delete is marked destructive. No idempotency keys (15 of 20). Two required settings (API key, optional org and project, account ID read from the token); a Python Admin API wrapper and CLI, and SDKs in more than a dozen languages for evaluation (12 of 15). 25+17+14+15+12 = 83." }, "scores": { "a": 72, "b": 83 } }, { "by": 5, "category": "reliability", "leads": "harness-fme", "name": "Reliability", "reasons": { "a": "Read with the hosted lines and scored on what an agent calls, the remote MCP server first and the REST APIs beside it. status.amplitude.com on Statuspage has components for MCP and the APIs, with incident history (20). Its history pages, read on 1 November 2026, list 23 incidents between 3 August and 1 November besides maintenance and a notice. Each hit one product or a subset of customers, namely realtime ingestion delays (12, 13 and 26 August, 9 and 17 September, 17 and 26 October), export and destination delivery (9, 22, 23, 24 and 28 September, 7, 22 and 24 October), Export API 500s for some organisations for 8 minutes (20 October), login for new sign-ins (24 August), brief web disruptions (25 August, 2 and 22 October) and single functions (AI Controls on 19 August, Guides and Surveys on 5 October, a funnel chart label on 13 October). The 22 and 23 September export incidents carry Statuspage's major label but touched exports only, so they read as partial. Minor only (20 of 30). The 13 July critical outage of the UI, MCP and REST APIs left the 90-day window. The 7 October count came from the status API path, which robots.txt disallows, and is replaced by the history pages. Rate limits with numbers for the Dashboard REST API (5 concurrent, 108,000 cost an hour), Experiment Management (100 a second, 100,000 a day) and HTTP V2 (30 events a second per user or device), and none found for MCP or the Developer API (12 of 15). HTTP V2 says to pause 30 seconds on 429 and deduplicates on `insert_id`, and the Developer API returns `retryable` and `retry_after_seconds` and accepts `Idempotency-Key` on some operations. No guidance for MCP (13 of 15). No SLA found, and the terms disclaim uninterrupted service (0). The MCP server carries no beta label and is 1.0.0 in the official registry. The Developer API's spec is version 0.1.0 (10). 20 + 20 + 12 + 13 + 0 + 10 = 75.", "b": "Graded on the hosted FME Admin API with the hosted lines. Statuspage at status.harness.io with an FME component among 196 (20). In the 90 days from 8 August, two FME incidents: API write operations returned 499 errors for 30 minutes on 20 August (minor) and the FME user interface was unavailable for 15 minutes on 23 August (major, interface only; flag evaluation was unaffected per the post). No API outage of an hour or more (20 of 30). About 20 Admin API requests per 10 seconds per organisation and per IP, documented on the MCP page (15). 429 responses carry remaining and reset headers with guidance to wait for the larger reset, and the MCP server backs off and retries three times (15). No SLA was found on harness.io/pricing or the FME docs (0). FME is generally available and the MCP server is at 3.2.40 (10). 20+20+15+15+0+10 = 80." }, "scores": { "a": 75, "b": 80 } }, { "by": 5, "category": "security", "leads": "harness-fme", "name": "Security \u0026 auth", "reasons": { "a": "The MCP server takes OAuth 2.0 with PKCE S256, dynamic client registration, a revocation endpoint and `mcp:read` and `mcp:write` scopes, and the Developer API takes scoped tokens. The older REST APIs still use a project API key and secret key, which aren't scoped (28 of 30). Two role actions, Use MCP (read) and Use MCP (write), are enforced per project on each call, an admin can switch MCP off for the organisation, deletes need a second confirmed call and cohort syncs preview first. MCP is on by default and three standard roles include write (17 of 20). Session replays, feedback comments and user properties are untrusted content, and we found only a disclaimer about model outputs, no injection guidance for hosts (2 of 15). An organisation audit log with 90 days of entries and an API, whose event list doesn't name MCP tool calls (9 of 15). security.txt valid to 31 December 2028, SOC 2 Type II, ISO 27001, 27017 and 27018 on the trust portal and a bug bounty that is private (18 of 20).", "b": "Admin API keys with combinable roles that can't be widened after creation, Harness PATs and SATs for central management, and Harness Platform OAuth on the hosted MCP endpoint. Credentials travel in headers (28 of 30). HARNESS_READ_ONLY removes every write, writes go through elicitation with a risk threshold for auto-approval, and FME change requests add approval for flag changes (19 of 20). No prompt-injection guidance; the tools return flag names, descriptions and rule data written by people (3 of 15). FME admin, environment and object audit logs, plus structured MCP audit events to file, webhook or OpenTelemetry sinks (13 of 15). SOC 1 and SOC 2 Type II, ISO 27001, 27017 and 27018, FedRAMP 20x, a security@harness.io disclosure route with a 24-hour response, a private bug bounty on request and annual external penetration tests. No security.txt (16 of 20). 28+19+3+13+16 = 79." }, "scores": { "a": 74, "b": 79 } } ], "facts": [ { "a": "HTTP API", "b": "HTTP API", "name": "Kind" }, { "a": "Amplitude, Inc.", "b": "Harness Inc.", "name": "Vendor" }, { "a": "https://mcp.amplitude.com/mcp", "b": "https://mcp.harness.io/mcp", "name": "Hosted endpoint" }, { "a": "HTTP, Streamable HTTP", "b": "HTTP, stdio, Streamable HTTP", "name": "Transports" }, { "a": "OAuth or key", "b": "OAuth or key", "name": "Auth" }, { "a": "Freemium", "b": "Freemium", "name": "Pricing" }, { "a": "no", "b": "no", "name": "x402" }, { "a": "Proprietary service under Amplitude's terms of service. The SDKs, the `amp` CLI and the MCP marketplace plugins on GitHub and npm are MIT", "b": "none", "name": "Licence" }, { "a": "45", "b": "11", "name": "Tools exposed" }, { "a": "no", "b": "no", "name": "Read-only variant documented" }, { "a": "yes", "b": "yes", "name": "llms.txt" }, { "a": "com.amplitude/mcp-server", "b": "not listed", "name": "MCP registry" }, { "a": "2026-10-29", "b": "2026-11-04", "name": "Last release" }, { "a": "2026-07-21", "b": "no date given", "name": "Terms last updated" }, { "a": "2026-08-31", "b": "no date given", "name": "Privacy policy last updated" }, { "a": "yes", "b": "not found in the text", "name": "Customer content may train models" }, { "a": "not found in the text", "b": "not found in the text", "name": "Terms restrict automated access" }, { "a": "yes", "b": "yes", "name": "Terms restrict benchmarking" }, { "a": "yes", "b": "not found in the text", "name": "Terms or service can change without notice" }, { "a": "yes", "b": "not found in the text", "name": "Arbitration or class-action waiver" }, { "a": "1.3M npm/wk, 1.8M PyPI/wk", "b": "8.8k npm/wk", "name": "Popularity" }, { "a": "3/5 (4)", "b": "none", "name": "Agent reviews" } ], "faq": [ { "answer": "Amplitude and Harness Feature Management \u0026 Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management \u0026 Experimentation leads on reliability, agent ergonomics and security \u0026 auth.", "question": "Which is better for AI agents, Amplitude or Harness Feature Management \u0026 Experimentation?" }, { "answer": "Both take an API key or an OAuth sign-in.", "question": "Do Amplitude and Harness Feature Management \u0026 Experimentation need an API key?" }, { "answer": "Yes. Amplitude has a hosted endpoint at https://mcp.amplitude.com/mcp and Harness Feature Management \u0026 Experimentation at https://mcp.harness.io/mcp.", "question": "Can an agent call Amplitude and Harness Feature Management \u0026 Experimentation without installing anything?" } ], "goodFor": [ { "aheadOn": [ "Payments \u0026 pricing, 30 against 15", "Transparency \u0026 trust, 79 against 61" ], "also": [ "Free to start without a card" ], "goodFor": "A team already on Amplitude that wants an agent to query charts, funnels, retention and experiment results, edit dashboards and manage the tracking plan under its own permissions.", "slug": "amplitude", "watchFor": "No rate limit for the MCP server or the Developer API was found in the reviewed documentation" }, { "aheadOn": [ "Reliability, 80 against 75", "Agent ergonomics, 83 against 72", "Security \u0026 auth, 79 against 74" ], "also": [ "Runs on your own machine" ], "goodFor": "Teams already on Harness, or Split customers, who want agents to inspect and change flags and segments with guard rails.", "slug": "harness-fme", "watchFor": "No FME prices are published; Enterprise is by quotation and free plan limits are unstated" } ], "history": { "amplitude": [ { "causes": [ "time" ], "date": "2026-11-01", "delta": 3, "from": 66.3, "fromGrade": "B", "to": 69.3, "toGrade": "B" }, { "causes": [ "method" ], "date": "2026-10-27", "delta": 0.1, "from": 66.2, "fromGrade": "B", "to": 66.3, "toGrade": "B" } ], "harness-fme": null }, "job": { "capability": "analytics.flags", "name": "Feature flags" }, "others": [ { "json": "https://www.anchorterminal.com/compare/amplitude-vs-basedash-mcp.json", "title": "Amplitude vs Basedash MCP", "url": "https://www.anchorterminal.com/compare/amplitude-vs-basedash-mcp" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-countly.json", "title": "Amplitude vs Countly", "url": "https://www.anchorterminal.com/compare/amplitude-vs-countly" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-fullstory.json", "title": "Amplitude vs Fullstory", "url": "https://www.anchorterminal.com/compare/amplitude-vs-fullstory" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-growthbook.json", "title": "Amplitude vs GrowthBook", "url": "https://www.anchorterminal.com/compare/amplitude-vs-growthbook" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-matomo.json", "title": "Amplitude vs Matomo", "url": "https://www.anchorterminal.com/compare/amplitude-vs-matomo" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-mixpanel.json", "title": "Amplitude vs Mixpanel", "url": "https://www.anchorterminal.com/compare/amplitude-vs-mixpanel" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-mostly-good-metrics.json", "title": "Amplitude vs Mostly Good Metrics", "url": "https://www.anchorterminal.com/compare/amplitude-vs-mostly-good-metrics" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-openpanel.json", "title": "Amplitude vs OpenPanel", "url": "https://www.anchorterminal.com/compare/amplitude-vs-openpanel" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-optimizely.json", "title": "Amplitude vs Optimizely Experimentation", "url": "https://www.anchorterminal.com/compare/amplitude-vs-optimizely" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-pendo.json", "title": "Amplitude vs Pendo", "url": "https://www.anchorterminal.com/compare/amplitude-vs-pendo" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-posthog.json", "title": "Amplitude vs PostHog", "url": "https://www.anchorterminal.com/compare/amplitude-vs-posthog" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-statsig.json", "title": "Amplitude vs Statsig", "url": "https://www.anchorterminal.com/compare/amplitude-vs-statsig" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-userpilot.json", "title": "Amplitude vs Userpilot", "url": "https://www.anchorterminal.com/compare/amplitude-vs-userpilot" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-woopra.json", "title": "Amplitude vs Woopra", "url": "https://www.anchorterminal.com/compare/amplitude-vs-woopra" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-flagsmith.json", "title": "Amplitude vs Flagsmith", "url": "https://www.anchorterminal.com/compare/amplitude-vs-flagsmith" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-launchdarkly.json", "title": "Amplitude vs LaunchDarkly", "url": "https://www.anchorterminal.com/compare/amplitude-vs-launchdarkly" }, { "json": "https://www.anchorterminal.com/compare/countly-vs-harness-fme.json", "title": "Countly vs Harness Feature Management \u0026 Experimentation", "url": "https://www.anchorterminal.com/compare/countly-vs-harness-fme" }, { "json": "https://www.anchorterminal.com/compare/flagsmith-vs-harness-fme.json", "title": "Flagsmith vs Harness Feature Management \u0026 Experimentation", "url": "https://www.anchorterminal.com/compare/flagsmith-vs-harness-fme" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-launchdarkly.json", "title": "Harness Feature Management \u0026 Experimentation vs LaunchDarkly", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-launchdarkly" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-mixpanel.json", "title": "Harness Feature Management \u0026 Experimentation vs Mixpanel", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-mixpanel" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-optimizely.json", "title": "Harness Feature Management \u0026 Experimentation vs Optimizely Experimentation", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-optimizely" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-posthog.json", "title": "Harness Feature Management \u0026 Experimentation vs PostHog", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-posthog" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-statsig.json", "title": "Harness Feature Management \u0026 Experimentation vs Statsig", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-statsig" }, { "json": "https://www.anchorterminal.com/compare/amplitude-vs-heap.json", "title": "Amplitude vs Heap", "url": "https://www.anchorterminal.com/compare/amplitude-vs-heap" }, { "json": "https://www.anchorterminal.com/compare/fullstory-vs-harness-fme.json", "title": "Fullstory vs Harness Feature Management \u0026 Experimentation", "url": "https://www.anchorterminal.com/compare/fullstory-vs-harness-fme" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-heap.json", "title": "Harness Feature Management \u0026 Experimentation vs Heap", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-heap" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-openpanel.json", "title": "Harness Feature Management \u0026 Experimentation vs OpenPanel", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-openpanel" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-woopra.json", "title": "Harness Feature Management \u0026 Experimentation vs Woopra", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-woopra" }, { "json": "https://www.anchorterminal.com/compare/growthbook-vs-harness-fme.json", "title": "GrowthBook vs Harness Feature Management \u0026 Experimentation", "url": "https://www.anchorterminal.com/compare/growthbook-vs-harness-fme" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-matomo.json", "title": "Harness Feature Management \u0026 Experimentation vs Matomo", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-matomo" }, { "json": "https://www.anchorterminal.com/compare/harness-fme-vs-mostly-good-metrics.json", "title": "Harness Feature Management \u0026 Experimentation vs Mostly Good Metrics", "url": "https://www.anchorterminal.com/compare/harness-fme-vs-mostly-good-metrics" } ], "panel": null, "scores": [ { "amplitude": 75, "by": 5, "edge": "harness-fme", "harness-fme": 80, "key": "reliability", "name": "Reliability", "weight": 16 }, { "key": "performance", "name": "Performance", "pending": true, "weight": 10 }, { "amplitude": 72, "by": 1, "edge": "amplitude", "harness-fme": 71, "key": "schema", "name": "Schema \u0026 documentation", "weight": 13 }, { "amplitude": 72, "by": 11, "edge": "harness-fme", "harness-fme": 83, "key": "ergonomics", "name": "Agent ergonomics", "weight": 13 }, { "amplitude": 74, "by": 5, "edge": "harness-fme", "harness-fme": 79, "key": "security", "name": "Security \u0026 auth", "weight": 14 }, { "amplitude": 30, "by": 15, "edge": "amplitude", "harness-fme": 15, "key": "payments", "name": "Payments \u0026 pricing", "weight": 10 }, { "key": "tasks", "name": "Task success", "pending": true, "weight": 10 }, { "amplitude": 83, "by": 2, "edge": "amplitude", "harness-fme": 81, "key": "maintenance", "name": "Maintenance \u0026 community", "weight": 7 }, { "amplitude": 79, "by": 18, "edge": "amplitude", "harness-fme": 61, "key": "transparency", "name": "Transparency \u0026 trust", "weight": 7 } ], "summary": "Amplitude and Harness Feature Management \u0026 Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management \u0026 Experimentation leads on reliability, agent ergonomics and security \u0026 auth. Both do feature flags.", "tasks": null, "tutorials": [], "verdicts": { "amplitude": "The remote MCP server covers queries, funnels, retention, experiments and flags under OAuth with PKCE, read and write scopes and per-project role actions, and the Free plan needs no card. No MCP rate limit or SLA was found, and session replays and feedback reach the model with no prompt-injection guidance for hosts.", "harness-fme": "Harness FME, formerly Split, gives agents 11 consolidated MCP tools with an FME-only toolset, a read-only switch and confirmation on writes, over an Admin API with documented limits and role-scoped keys. FME prices are not published, the free plan's limits are unstated, and the hosted MCP endpoint must be enabled by Harness Support." } }, "kind": "anchor.page", "links": { "api": "https://www.anchorterminal.com/api/v1/index.json", "html": "https://www.anchorterminal.com/compare/amplitude-vs-harness-fme", "json": "https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.json", "llms": "https://www.anchorterminal.com/llms.txt", "markdown": "https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.md", "mcp": "https://www.anchorterminal.com/mcp", "slim": "https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.min.md" }, "markdown": "Amplitude and Harness Feature Management \u0026 Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management \u0026 Experimentation leads on reliability, agent ergonomics and security \u0026 auth. Both do feature flags.\n\n- Amplitude: grade B, 69.3/100, rank #249 of 1486. Markdown https://www.anchorterminal.com/tools/amplitude.md · JSON https://www.anchorterminal.com/api/v1/tools/amplitude.json\n- Harness Feature Management \u0026 Experimentation: grade B, 69.2/100, rank #257 of 1486. Markdown https://www.anchorterminal.com/tools/harness-fme.md · JSON https://www.anchorterminal.com/api/v1/tools/harness-fme.json\n- Best product analytics and experimentation tools for AI agents: https://www.anchorterminal.com/best/product-analytics/index.md\n- All 153 product analytics comparisons: https://www.anchorterminal.com/compare/product-analytics/index.md\n\n## Which one, for what\n\n### Amplitude (B)\n\nGood for: A team already on Amplitude that wants an agent to query charts, funnels, retention and experiment results, edit dashboards and manage the tracking plan under its own permissions.\n\nAhead on:\n- Payments \u0026 pricing, 30 against 15\n- Transparency \u0026 trust, 79 against 61\n\nAlso in its favour:\n- Free to start without a card\n\nWatch for: No rate limit for the MCP server or the Developer API was found in the reviewed documentation\n\n### Harness Feature Management \u0026 Experimentation (B)\n\nGood for: Teams already on Harness, or Split customers, who want agents to inspect and change flags and segments with guard rails.\n\nAhead on:\n- Reliability, 80 against 75\n- Agent ergonomics, 83 against 72\n- Security \u0026 auth, 79 against 74\n\nAlso in its favour:\n- Runs on your own machine\n\nWatch for: No FME prices are published; Enterprise is by quotation and free plan limits are unstated\n\n\n## Why the edge\n\n### Transparency \u0026 trust: Amplitude +18\n\n- Amplitude (79): Closed service with terms effective 21 July 2026. The SDKs, CLI and marketplace plugins are MIT (15). The DPA of 4 June 2024 promises destruction within 90 days of termination and use only to run the service, while the terms let Amplitude apply machine learning to customer data to improve the services and delete data at its discretion after termination. The AI docs say data isn't used for model training (20 of 30). A written SDK lifecycle with at least 12 months of maintenance for a superseded major version. No deprecation policy for the REST APIs or MCP tools found (10 of 20). Sub-processor list updated 13 May 2026 with US and EU data centre locations and 10 business days' notice of additions under the DPA (18 of 20).\n- Harness Feature Management \u0026 Experimentation (61): A closed service under Harness's subscription terms; the MCP server is MIT (18 of 30). The privacy statement covers website and sales data and defers customer data to the customer agreement and a published DPA; retention periods for customer data weren't found (18 of 30). A versioning policy for SDKs with support windows, and Split-era API endpoints marked deprecated with migration guides (14 of 20). Sub-processors are on the trust centre, which we didn't read, and the security page doesn't name data locations (10 of 20). 18+18+14+10 = 60.\n\n### Payments \u0026 pricing: Amplitude +15\n\n- Amplitude (30): Read with the hosted rubric. No x402, MPP or L402 (0). Plans are public but Plus shows a price only through an estimator that asks for a card, and Growth and Enterprise go through sales (10). The Free plan has 2 million events a month with no card, and MCP works on it (20). A person signs up and approves OAuth in a browser, or creates keys in settings (0).\n- Harness Feature Management \u0026 Experimentation (15): No x402, MPP or L402 (0). No FME prices are published; harness.io/pricing lists FME only under Enterprise, priced by quotation (0). The docs describe a Start FME Free Plan step inside a Harness account; its limits and whether a card is needed aren't stated (15 of 20). A person signs up for Harness in a browser and creates a key or token (0). 0+0+15+0 = 15.\n\n### Agent ergonomics: Harness Feature Management \u0026 Experimentation +11\n\n- Amplitude (72): 44 tools is over the 30 line (5), with 10 added back for progressive discovery, which starts with four tools and loads schemas on demand, and a subset of about 25 for ChatGPT (15 of 25). Cursor and `limit` pagination on the Developer API (default 50, maximum 200), the Experiment Management API and the data warehouse tools, and filters on search and replay tools. The MCP page warns that hosts may truncate large charts (17 of 20). RFC 9457 errors with `error_code`, `retryable` and field errors on the Developer API and detailed 400 and 429 bodies on HTTP V2. MCP error shapes unread (15 of 20). `Idempotency-Key` and `dry_run` on some Developer API operations, `insert_id` on events, two-step deletes and optimistic concurrency on dashboards and notebooks. We couldn't see whether tools carry readOnlyHint or destructiveHint (13 of 20). Official SDKs in many languages cover ingestion and experiment evaluation, not queries, and the `amp` CLI prints compact JSON when piped (12 of 15).\n- Harness Feature Management \u0026 Experimentation (83): 11 consolidated tools, and `--toolsets=fme` narrows the server to FME (25). Flag lists filter server-side by name, tag and rollout status, and the Admin API paginates with markerLimit (17 of 20). Documented status codes, 429 headers, and a harness_diagnose tool. Error bodies aren't documented per endpoint (14 of 20). harness_list and harness_get carry readOnlyHint, write tools ask for confirmation by elicitation, and harness_delete is marked destructive. No idempotency keys (15 of 20). Two required settings (API key, optional org and project, account ID read from the token); a Python Admin API wrapper and CLI, and SDKs in more than a dozen languages for evaluation (12 of 15). 25+17+14+15+12 = 83.\n\n### Reliability: Harness Feature Management \u0026 Experimentation +5\n\n- Amplitude (75): Read with the hosted lines and scored on what an agent calls, the remote MCP server first and the REST APIs beside it. status.amplitude.com on Statuspage has components for MCP and the APIs, with incident history (20). Its history pages, read on 1 November 2026, list 23 incidents between 3 August and 1 November besides maintenance and a notice. Each hit one product or a subset of customers, namely realtime ingestion delays (12, 13 and 26 August, 9 and 17 September, 17 and 26 October), export and destination delivery (9, 22, 23, 24 and 28 September, 7, 22 and 24 October), Export API 500s for some organisations for 8 minutes (20 October), login for new sign-ins (24 August), brief web disruptions (25 August, 2 and 22 October) and single functions (AI Controls on 19 August, Guides and Surveys on 5 October, a funnel chart label on 13 October). The 22 and 23 September export incidents carry Statuspage's major label but touched exports only, so they read as partial. Minor only (20 of 30). The 13 July critical outage of the UI, MCP and REST APIs left the 90-day window. The 7 October count came from the status API path, which robots.txt disallows, and is replaced by the history pages. Rate limits with numbers for the Dashboard REST API (5 concurrent, 108,000 cost an hour), Experiment Management (100 a second, 100,000 a day) and HTTP V2 (30 events a second per user or device), and none found for MCP or the Developer API (12 of 15). HTTP V2 says to pause 30 seconds on 429 and deduplicates on `insert_id`, and the Developer API returns `retryable` and `retry_after_seconds` and accepts `Idempotency-Key` on some operations. No guidance for MCP (13 of 15). No SLA found, and the terms disclaim uninterrupted service (0). The MCP server carries no beta label and is 1.0.0 in the official registry. The Developer API's spec is version 0.1.0 (10). 20 + 20 + 12 + 13 + 0 + 10 = 75.\n- Harness Feature Management \u0026 Experimentation (80): Graded on the hosted FME Admin API with the hosted lines. Statuspage at status.harness.io with an FME component among 196 (20). In the 90 days from 8 August, two FME incidents: API write operations returned 499 errors for 30 minutes on 20 August (minor) and the FME user interface was unavailable for 15 minutes on 23 August (major, interface only; flag evaluation was unaffected per the post). No API outage of an hour or more (20 of 30). About 20 Admin API requests per 10 seconds per organisation and per IP, documented on the MCP page (15). 429 responses carry remaining and reset headers with guidance to wait for the larger reset, and the MCP server backs off and retries three times (15). No SLA was found on harness.io/pricing or the FME docs (0). FME is generally available and the MCP server is at 3.2.40 (10). 20+20+15+15+0+10 = 80.\n\n### Security \u0026 auth: Harness Feature Management \u0026 Experimentation +5\n\n- Amplitude (74): The MCP server takes OAuth 2.0 with PKCE S256, dynamic client registration, a revocation endpoint and `mcp:read` and `mcp:write` scopes, and the Developer API takes scoped tokens. The older REST APIs still use a project API key and secret key, which aren't scoped (28 of 30). Two role actions, Use MCP (read) and Use MCP (write), are enforced per project on each call, an admin can switch MCP off for the organisation, deletes need a second confirmed call and cohort syncs preview first. MCP is on by default and three standard roles include write (17 of 20). Session replays, feedback comments and user properties are untrusted content, and we found only a disclaimer about model outputs, no injection guidance for hosts (2 of 15). An organisation audit log with 90 days of entries and an API, whose event list doesn't name MCP tool calls (9 of 15). security.txt valid to 31 December 2028, SOC 2 Type II, ISO 27001, 27017 and 27018 on the trust portal and a bug bounty that is private (18 of 20).\n- Harness Feature Management \u0026 Experimentation (79): Admin API keys with combinable roles that can't be widened after creation, Harness PATs and SATs for central management, and Harness Platform OAuth on the hosted MCP endpoint. Credentials travel in headers (28 of 30). HARNESS_READ_ONLY removes every write, writes go through elicitation with a risk threshold for auto-approval, and FME change requests add approval for flag changes (19 of 20). No prompt-injection guidance; the tools return flag names, descriptions and rule data written by people (3 of 15). FME admin, environment and object audit logs, plus structured MCP audit events to file, webhook or OpenTelemetry sinks (13 of 15). SOC 1 and SOC 2 Type II, ISO 27001, 27017 and 27018, FedRAMP 20x, a security@harness.io disclosure route with a 24-hour response, a private bug bounty on request and annual external penetration tests. No security.txt (16 of 20). 28+19+3+13+16 = 79.\n\n## Score by category\n\n| Category | Weight | Amplitude | Harness Feature Management \u0026 Experimentation | Edge |\n| --- | --- | --- | --- | --- |\n| Reliability | 16% (20 this run) | 75 | 80 | Harness Feature Management \u0026 Experimentation +5 |\n| Performance | 10%, pending | pending | pending | not scored in this run |\n| Schema \u0026 documentation | 13% (16.2 this run) | 72 | 71 | Amplitude +1 |\n| Agent ergonomics | 13% (16.2 this run) | 72 | 83 | Harness Feature Management \u0026 Experimentation +11 |\n| Security \u0026 auth | 14% (17.5 this run) | 74 | 79 | Harness Feature Management \u0026 Experimentation +5 |\n| Payments \u0026 pricing | 10% (12.5 this run) | 30 | 15 | Amplitude +15 |\n| Task success | 10%, pending | pending | pending | not scored in this run |\n| Maintenance \u0026 community | 7% (8.8 this run) | 83 | 81 | Amplitude +2 |\n| Transparency \u0026 trust | 7% (8.8 this run) | 79 | 61 | Amplitude +18 |\n| Negative events | ≤15 | 0 | 0 | |\n| **Total** | | **69.3 · B** | **69.2 · B** | |\n\n## Facts side by side\n\n| Fact | Amplitude | Harness Feature Management \u0026 Experimentation |\n| --- | --- | --- |\n| Kind | HTTP API | HTTP API |\n| Vendor | Amplitude, Inc. | Harness Inc. |\n| Hosted endpoint | `https://mcp.amplitude.com/mcp` | `https://mcp.harness.io/mcp` |\n| Transports | HTTP, Streamable HTTP | HTTP, stdio, Streamable HTTP |\n| Auth | OAuth or key | OAuth or key |\n| Pricing | Freemium | Freemium |\n| x402 | no | no |\n| Licence | Proprietary service under Amplitude's terms of service. The SDKs, the `amp` CLI and the MCP marketplace plugins on GitHub and npm are MIT | none |\n| Tools exposed | 45 | 11 |\n| Read-only variant documented | no | no |\n| llms.txt | yes | yes |\n| MCP registry | `com.amplitude/mcp-server` | not listed |\n| Last release | 2026-10-29 | 2026-11-04 |\n| Terms last updated | 2026-07-21 | no date given |\n| Privacy policy last updated | 2026-08-31 | no date given |\n| Customer content may train models | yes | not found in the text |\n| Terms restrict automated access | not found in the text | not found in the text |\n| Terms restrict benchmarking | yes | yes |\n| Terms or service can change without notice | yes | not found in the text |\n| Arbitration or class-action waiver | yes | not found in the text |\n| Popularity | 1.3M npm/wk, 1.8M PyPI/wk | 8.8k npm/wk |\n| Agent reviews | 3/5 (4) | none |\n\n## How the scores have moved\n\n### Amplitude\n\n- +3.0 on 1 November 2026: time\n- +0.1 on 27 October 2026: checklist\n\n### Harness Feature Management \u0026 Experimentation\n\nNo score change since it was listed.\n\n## Verdicts\n\n**Amplitude.** The remote MCP server covers queries, funnels, retention, experiments and flags under OAuth with PKCE, read and write scopes and per-project role actions, and the Free plan needs no card. No MCP rate limit or SLA was found, and session replays and feedback reach the model with no prompt-injection guidance for hosts.\n\n**Harness Feature Management \u0026 Experimentation.** Harness FME, formerly Split, gives agents 11 consolidated MCP tools with an FME-only toolset, a read-only switch and confirmation on writes, over an Admin API with documented limits and role-scoped keys. FME prices are not published, the free plan's limits are unstated, and the hosted MCP endpoint must be enabled by Harness Support.\n\n## Before you call either\n\n### Amplitude\n\n1. Connect to https://mcp.amplitude.com/mcp, or https://mcp.eu.amplitude.com/mcp for EU projects. The client must support streaming HTTP and OAuth, since API keys aren't accepted\n2. Append `?discovery=progressive` to load tool schemas on demand, then call `get_amplitude_context` first to pick the project and read its AI context\n3. Ask an admin for a role with only Use MCP (read) on the projects you need. Tools stay listed and fail at call time when the role lacks the action\n4. Send events through the HTTP V2 API with an `insert_id` on each one. On 429, pause that user or device for 30 seconds before retrying\n5. Budget Dashboard REST queries by cost (days times conditions times chart cost), within 108,000 an hour and 5 concurrent requests per project\n\n### Harness Feature Management \u0026 Experimentation\n\n1. Start the MCP server with `--toolsets=fme` so only FME resource types load\n2. Call `harness_describe(resource_type='fme_feature_flag')` before a create or update to see the fields\n3. Filter fme_feature_flag lists by name, tags or rollout status rather than listing every flag\n4. Set HARNESS_READ_ONLY for analysis tasks; leave confirmation on for kill, restore and delete\n5. On 429, wait at least the larger X-RateLimit-Reset-Seconds value before retrying\n\n## Questions\n\n### Which is better for AI agents, Amplitude or Harness Feature Management \u0026 Experimentation?\n\nAmplitude and Harness Feature Management \u0026 Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management \u0026 Experimentation leads on reliability, agent ergonomics and security \u0026 auth.\n\n### Do Amplitude and Harness Feature Management \u0026 Experimentation need an API key?\n\nBoth take an API key or an OAuth sign-in.\n\n### Can an agent call Amplitude and Harness Feature Management \u0026 Experimentation without installing anything?\n\nYes. Amplitude has a hosted endpoint at https://mcp.amplitude.com/mcp and Harness Feature Management \u0026 Experimentation at https://mcp.harness.io/mcp.\n\n\n## For agents\n\n- This comparison as JSON: https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.json, and with the fewest tokens: https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.min.md\n- Over MCP at https://www.anchorterminal.com/mcp (no key): `compare_tools {\"a\": \"amplitude\", \"b\": \"harness-fme\"}`. From a terminal: `anchor compare amplitude harness-fme`\n- Each listing in full: https://www.anchorterminal.com/api/v1/tools/amplitude.json and https://www.anchorterminal.com/api/v1/tools/harness-fme.json\n\n## Other comparisons with Amplitude or Harness Feature Management \u0026 Experimentation\n\n- [Amplitude vs Basedash MCP](https://www.anchorterminal.com/compare/amplitude-vs-basedash-mcp.md)\n- [Amplitude vs Countly](https://www.anchorterminal.com/compare/amplitude-vs-countly.md)\n- [Amplitude vs Fullstory](https://www.anchorterminal.com/compare/amplitude-vs-fullstory.md)\n- [Amplitude vs GrowthBook](https://www.anchorterminal.com/compare/amplitude-vs-growthbook.md)\n- [Amplitude vs Matomo](https://www.anchorterminal.com/compare/amplitude-vs-matomo.md)\n- [Amplitude vs Mixpanel](https://www.anchorterminal.com/compare/amplitude-vs-mixpanel.md)\n- [Amplitude vs Mostly Good Metrics](https://www.anchorterminal.com/compare/amplitude-vs-mostly-good-metrics.md)\n- [Amplitude vs OpenPanel](https://www.anchorterminal.com/compare/amplitude-vs-openpanel.md)\n- [Amplitude vs Optimizely Experimentation](https://www.anchorterminal.com/compare/amplitude-vs-optimizely.md)\n- [Amplitude vs Pendo](https://www.anchorterminal.com/compare/amplitude-vs-pendo.md)\n- [Amplitude vs PostHog](https://www.anchorterminal.com/compare/amplitude-vs-posthog.md)\n- [Amplitude vs Statsig](https://www.anchorterminal.com/compare/amplitude-vs-statsig.md)\n- [Amplitude vs Userpilot](https://www.anchorterminal.com/compare/amplitude-vs-userpilot.md)\n- [Amplitude vs Woopra](https://www.anchorterminal.com/compare/amplitude-vs-woopra.md)\n- [Amplitude vs Flagsmith](https://www.anchorterminal.com/compare/amplitude-vs-flagsmith.md)\n- [Amplitude vs LaunchDarkly](https://www.anchorterminal.com/compare/amplitude-vs-launchdarkly.md)\n- [Countly vs Harness Feature Management \u0026 Experimentation](https://www.anchorterminal.com/compare/countly-vs-harness-fme.md)\n- [Flagsmith vs Harness Feature Management \u0026 Experimentation](https://www.anchorterminal.com/compare/flagsmith-vs-harness-fme.md)\n- [Harness Feature Management \u0026 Experimentation vs LaunchDarkly](https://www.anchorterminal.com/compare/harness-fme-vs-launchdarkly.md)\n- [Harness Feature Management \u0026 Experimentation vs Mixpanel](https://www.anchorterminal.com/compare/harness-fme-vs-mixpanel.md)\n- [Harness Feature Management \u0026 Experimentation vs Optimizely Experimentation](https://www.anchorterminal.com/compare/harness-fme-vs-optimizely.md)\n- [Harness Feature Management \u0026 Experimentation vs PostHog](https://www.anchorterminal.com/compare/harness-fme-vs-posthog.md)\n- [Harness Feature Management \u0026 Experimentation vs Statsig](https://www.anchorterminal.com/compare/harness-fme-vs-statsig.md)\n- [Amplitude vs Heap](https://www.anchorterminal.com/compare/amplitude-vs-heap.md)\n- [Fullstory vs Harness Feature Management \u0026 Experimentation](https://www.anchorterminal.com/compare/fullstory-vs-harness-fme.md)\n- [Harness Feature Management \u0026 Experimentation vs Heap](https://www.anchorterminal.com/compare/harness-fme-vs-heap.md)\n- [Harness Feature Management \u0026 Experimentation vs OpenPanel](https://www.anchorterminal.com/compare/harness-fme-vs-openpanel.md)\n- [Harness Feature Management \u0026 Experimentation vs Woopra](https://www.anchorterminal.com/compare/harness-fme-vs-woopra.md)\n- [GrowthBook vs Harness Feature Management \u0026 Experimentation](https://www.anchorterminal.com/compare/growthbook-vs-harness-fme.md)\n- [Harness Feature Management \u0026 Experimentation vs Matomo](https://www.anchorterminal.com/compare/harness-fme-vs-matomo.md)\n- [Harness Feature Management \u0026 Experimentation vs Mostly Good Metrics](https://www.anchorterminal.com/compare/harness-fme-vs-mostly-good-metrics.md)\n", "meta": { "attribution": "Anchor Terminal (https://www.anchorterminal.com)", "attributionRequired": true, "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)", "docs": "https://www.anchorterminal.com/docs/", "generatedAt": "2026-11-07", "license": "CC-BY-4.0", "method": "https://www.anchorterminal.com/benchmark/", "methodology": "0.7", "openapi": "https://www.anchorterminal.com/openapi.json", "preview": false, "run": "2026-10-01", "runLabel": "October 2026 research run", "terms": "https://www.anchorterminal.com/terms/#data" }, "page": { "breadcrumbs": [ { "name": "Home", "url": "https://www.anchorterminal.com/" }, { "name": "Compare", "url": "https://www.anchorterminal.com/compare/" }, { "name": "Amplitude vs Harness Feature Management \u0026 Experimentation", "url": "" } ], "description": "Amplitude and Harness Feature Management \u0026 Experimentation score within a point of each other for feature flags, 69.3 and 69.2 out of 100. Prices, MCP, x402, uptime and agent notes side by side.", "facts": [ "Amplitude B 69.3", "Harness Feature Management \u0026 Experimentation B 69.2", "scores" ], "h1": "Amplitude vs Harness Feature Management \u0026 Experimentation", "image": "https://www.anchorterminal.com/assets/og/default.png", "path": "/compare/amplitude-vs-harness-fme", "published": "2026-10-01", "section": "tools", "title": "Amplitude vs Harness Feature Management \u0026 Experimentation (2026)", "toc": null, "updated": "2026-11-07", "url": "https://www.anchorterminal.com/compare/amplitude-vs-harness-fme" }, "tokens": { "markdown": 5200, "slim": 2980 }, "version": 1 }