# Amplitude vs Harness Feature Management & Experimentation (slim)

> Amplitude and Harness Feature Management & 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.

- Full: https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.md (~5,200 tokens) · this version ~2,980 tokens · JSON https://www.anchorterminal.com/compare/amplitude-vs-harness-fme.json · canonical https://www.anchorterminal.com/compare/amplitude-vs-harness-fme
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-07
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Amplitude and Harness Feature Management & Experimentation score within a point of each other on agent readiness, 69.3 (B) and 69.2 (B). Harness Feature Management & Experimentation leads on reliability, agent ergonomics and security & auth. Both do feature flags.

- Amplitude: B 69.3, rank #249 of 1486 · https://www.anchorterminal.com/tools/amplitude.min.md
- Harness Feature Management & Experimentation: B 69.2, rank #257 of 1486 · https://www.anchorterminal.com/tools/harness-fme.min.md
- Amplitude, good 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. Ahead on Payments & pricing, 30 against 15; Transparency & trust, 79 against 61. Also Free to start without a card.
- Harness Feature Management & Experimentation, good for Teams already on Harness, or Split customers, who want agents to inspect and change flags and segments with guard rails. Ahead on Reliability, 80 against 75; Agent ergonomics, 83 against 72; Security & auth, 79 against 74. Also Runs on your own machine.

| Category | Amplitude | Harness Feature Management & Experimentation |
| --- | --- | --- |
| Reliability (16%) | 75 | 80 |
| Performance (10%) | pending | pending |
| Schema & documentation (13%) | 72 | 71 |
| Agent ergonomics (13%) | 72 | 83 |
| Security & auth (14%) | 74 | 79 |
| Payments & pricing (10%) | 30 | 15 |
| Task success (10%) | pending | pending |
| Maintenance & community (7%) | 83 | 81 |
| Transparency & trust (7%) | 79 | 61 |

| Fact (where they differ) | Amplitude | Harness Feature Management & Experimentation |
| --- | --- | --- |
| Vendor | Amplitude, Inc. | Harness Inc. |
| Hosted endpoint | `https://mcp.amplitude.com/mcp` | `https://mcp.harness.io/mcp` |
| Transports | HTTP, Streamable HTTP | HTTP, stdio, Streamable HTTP |
| 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 |
| Tools exposed | 45 | 11 |
| MCP registry | `com.amplitude/mcp-server` | not listed |
| Last release | 2026-10-29 | 2026-11-04 |
| Terms last updated | 2026-07-21 | no date given |
| Privacy policy last updated | 2026-08-31 | no date given |
| Customer content may train models | yes | not found in the text |
| Terms or service can change without notice | yes | not found in the text |
| Arbitration or class-action waiver | yes | not found in the text |
| Popularity | 1.3M npm/wk, 1.8M PyPI/wk | 8.8k npm/wk |
| Agent reviews | 3/5 (4) | none |

Why the edge:
- Transparency & trust, Amplitude +18: 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).); Harness Feature Management & 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.)
- Payments & pricing, Amplitude +15: 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).); Harness Feature Management & 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.)
- Agent ergonomics, Harness Feature Management & Experimentation +11: 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).); Harness Feature Management & 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.)
- Reliability, Harness Feature Management & Experimentation +5: 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.); Harness Feature Management & 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.)
- Security & auth, Harness Feature Management & Experimentation +5: 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).); Harness Feature Management & 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.)