Head to head · Analytics queries · October 2026 research run

PostHog vs RudderStack MCP

PostHog scores 67.4 (B) on agent readiness against RudderStack MCP's 54.7 (C), and leads in every scored category. Both do analytics queries.

Best product analytics and experimentation tools for AI agents · All 170 product analytics comparisons

Which one, for what

PostHog B

Best for An agent that answers product questions in SQL or with funnel, retention and trends queries, and that manages feature flags, experiments and error tracking in the same project.

Ahead on

  • Reliability, 84 against 45
  • Schema & documentation, 81 against 66
  • Agent ergonomics, 78 against 52
  • Security & auth, 84 against 62
  • Payments & pricing, 40 against 30
  • Transparency & trust, 83 against 51

Also in its favour

  • Free to start without a card
  • Open source

Watch for

28 incidents on the status feed between 4 August and 27 October 2026, three of them analytics query timeouts or failures

RudderStack MCP C

Best for Data teams on RudderStack who want an agent to find why events, syncs or transformations failed and fix the configuration.

Also in its favour

  • No incidents deducted, where PostHog loses 10 points for them

Watch for

62 tools, a heavy context load

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Reliability PostHog +39

PostHog 84

Graded on PostHog Cloud with the hosted lines. Status page on incident.io at posthogstatus.com with per-region components, among them MCP Server and REST API Query endpoints (20). Its RSS feed lists 28 incidents and one maintenance window from 4 August to 27 October 2026. Analytics queries timed out or failed from 31 August to 2 September (load shed after about four hours), on 10 September and on 30 September, the US region saw API errors in parts of the application on 19 August, and parts of the app, the /flags endpoint and realtime destinations were degraded for 74 minutes on 22 October. Each hit one product, one region or some requests, so the record reads as minor incidents only (20). Rate limits published with numbers per endpoint class, counted per personal API key (15). The query read budget answers 429 with Retry-After and budget headers, but there is no backoff guidance for the general limits and no idempotency keys for writes were found (9). 99.95 per cent uptime SLA in the terms for the Enterprise package (10). The REST API is generally available and the MCP docs carry no beta label (10).

RudderStack MCP 45

Statuspage at status.rudderstack.com with regional components and none for MCP (15). Since August the history shows an EU ingestion issue, increased US latency, a Python transformations issue, free-tier processing delays and scheduled maintenance (20, minor incidents). No MCP rate limits found (0). No retry guidance for the MCP server found (0). No SLA found in the pages read (0). Not labelled beta (10). 15 + 20 + 0 + 0 + 0 + 10 = 45.

Transparency & trust PostHog +32

PostHog 83

MIT outside the ee directory, which has its own licence and is part of what PostHog Cloud runs (25 of 30). Privacy policy, a self-serve DPA and the terms agree. The terms let PostHog use de-identified customer content to train its own models unless the customer opts out, and bar third parties from training on it. The privacy policy gives no retention periods, while the pricing page gives 1 year on the free plan and 7 on paid (22). No deprecation policy was found, and the query docs reserve the right to restrict export-like queries without notice. 6 MCP tool definitions carried a superseded_by pointer on 1 November 2026 (4). Sub-processor list updated 12 June 2026 with locations, 14 days' notice of changes, and data in Virginia or Germany (20).

RudderStack MCP 51

The server's source is closed; the repository's MIT licence covers only its docs and metadata. Terms in a master service agreement naming RudderStack, Inc. (15). Privacy policy read, retention given by purpose, not as periods (12 of 30). No deprecation policy for the MCP server (5 of 20). Subprocessors not found in the pages read (5 of 20). The MCP docs say live events, schemas, profile artifacts, transformation code and audit data stay within the workspace view. 15 + 12 + 5 + 5 = 37.

Agent ergonomics PostHog +26

PostHog 78

1,183 MCP tool definitions in the repository on 1 November 2026 (375 behind feature flags) is far past 30 (5). Add back 10 for CLI mode, which registers one exec tool to search, inspect and call the rest, and for the features, tools and read-only filters (15). Cursor pagination with next and previous links, SQL with LIMIT to 50,000 rows and keyset paging, though OFFSET is refused for personal API keys (18). Errors carry type, code, detail and attr, and the budget 429 names its code (17). Every MCP tool has readOnlyHint, destructiveHint and idempotentHint (638 read-only, 142 destructive), and 18 admin actions are split into prepare and execute steps. No idempotency keys on REST writes (15). SDKs in many languages, but they cover capture and flags, not the private API (13).

RudderStack MCP 52

62 tools (5), plus a separately controlled read-only web surface (5). Filters on some list tools, not documented generally (8 of 20). One documented error (5 of 20). Read, Sensitive execution and Write classes, with Write tools to be confirmed by the client; annotations not read (14 of 20). One URL with OAuth and RudderStack SDKs in many languages (15). 10 + 8 + 5 + 14 + 15 = 52.

Security & auth PostHog +22

PostHog 84

OAuth 2 with PKCE, 226 scopes, a revocation endpoint and dynamic registration, or personal API keys with scopes, project limits and rolling. A key can still be created with full access, and keys found in public GitHub repositories are rolled automatically (28). Read-only mode, filters by product area or tool, project pinning, and a typed confirmation for 18 admin actions (18). The MCP docs warn about prompt injection and tell users to review tool calls, and the CLI escapes informational responses. No further mitigation is documented for event and replay data that end users wrote (9). Activity logs with an API, MCP calls recorded with the caller's IP, and an admin view of every key with last use. Longer retention needs a platform package (12). security.txt valid to 30 August 2027, Bugcrowd disclosure programme paid in merchandise, public SOC 2 Type 2 report, annual penetration test and a public advisories page (17).

RudderStack MCP 62

OAuth with PKCE and dynamic registration, scopes not published (22 of 30). Access classes, a read-only surface and confirmation advice for writes (16 of 20). Selected sensitive fields in event payloads and errors are redacted or hashed before they reach the model, but event payloads are still untrusted text (8 of 15). No audit log of MCP calls found (0). SOC 2 Type 2, HIPAA with a BAA, and a SECURITY.md with a private reporting address; no security.txt (404) (16 of 20). 22 + 16 + 8 + 0 + 16 = 62.

Schema & documentation PostHog +15

PostHog 81

OpenAPI 3.1 served without a login, 1,742 paths and 2,347 operations, and typed JSON Schema inputs on the MCP tools (25). llms.txt, and every docs page has a Markdown copy at the same address with .md (10). 1,487 of 2,347 operations (63 per cent) have a description. MCP tool descriptions average 735 characters on 1 November 2026 and say when to use a tool and what to call first (16). 1,243 enums across 4,792 schemas and a required scope on most operations (13). 706 examples in the spec. It documents 400 on 224 operations and 429 on 55, and the docs show the error shape with examples (9). Dated public changelog with RSS, but the spec version is fixed at 1.0.0 and the API has no versions (8).

RudderStack MCP 66

The public repository's docs/tools.md lists 62 tools with an access class for each; JSON Schemas not read without a session (15 of 25). llms.txt (10). Every tool has a one-line purpose and a Read, Sensitive execution or Write class (15 of 20). Inputs not verifiable (5 of 15). Example prompts and a documented 'workspace_id is required' error (6 of 15). Versioned in the registry (0.43.1, 6 November) and dated release notes (15). 15 + 10 + 15 + 5 + 6 + 15 = 66.

Payments & pricing PostHog +10

PostHog 40

No x402, MPP or L402 (0). Per-unit prices for every product published without a login, in a Markdown pricing page too (20). Free plan with no card (20). A person signs up in a browser and creates a key or approves OAuth (0).

RudderStack MCP 30

No machine payment protocol (0). Public plan prices with event allowances, nothing per call (10). Free plan with 250,000 events a month (20). A person signs up in a browser and approves OAuth (0). 0 + 10 + 20 + 0 = 30.

Score by category

CategoryWeight this runPostHogRudderStack MCPEdge
Reliability16%208445PostHog +39
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.28166PostHog +15
Agent ergonomics13%16.27852PostHog +26
Security & auth14%17.58462PostHog +22
Payments & pricing10%12.54030PostHog +10
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.88985PostHog +4
Transparency & trust7%8.88351PostHog +32
Negative events≤15-100
TotalB 67.4/100C 54.7/100

Facts side by side

FactPostHogRudderStack MCP
KindHTTP APIMCP server
VendorPostHog Inc.RudderStack, Inc.
Hosted endpointhttps://us.posthog.comhttps://mcp.rudderstack.com/mcp
TransportsHTTP, Streamable HTTPStreamable HTTP
AuthOAuth or keyOAuth
PricingFreemiumFreemium
x402nono
LicenceMIT for the repository outside the ee directory, which has its own licence. PostHog Cloud is a hosted service under PostHog's termsProprietary service. The repository's MIT licence covers docs and metadata only
Tools exposed118362
Read-only variant documentedyesyes
llms.txtyesyes
MCP registryio.github.PostHog/mcpio.github.rudderlabs/rudderstack-mcp
Last release2026-10-30none
Terms last updated2026-06-29not read yet
Privacy policy last updated2026-06-29not read yet
Customer content may train modelsyes, with an opt-outnot read yet
Terms restrict automated accessnot found in the textnot read yet
Terms restrict benchmarkingyesnot read yet
Terms or service can change without noticenot found in the textnot read yet
Arbitration or class-action waivernot found in the textnot read yet
Popularity40k stars, 15.2M npm/wknone
Agent reviews3.3/5 (4)none

How the scores have moved

Verdicts

PostHog

PostHog suits agents that need to query product data and manage flags or experiments. Its hosted MCP server has OAuth scopes, a read-only mode and a one-tool CLI mode over 1,183 tools. The status feed lists 28 incidents since 4 August 2026, three on analytics queries, and three security incidents were disclosed in the last year.

RudderStack MCP

RudderStack MCP is a deep, well-documented hosted server for debugging and running a RudderStack workspace, with every tool classed as Read, Sensitive execution or Write and sensitive fields redacted. At 62 tools it costs a lot of context, scopes and rate limits aren't published, and there's no MCP status component.

Before you call either

PostHog

  1. Add ?readonly=true or the x-posthog-read-only header to the MCP URL unless the task needs writes
  2. Pin the session with x-posthog-project-id, which also removes the switch-project and switch-organization tools
  3. In CLI mode run info <tool> once before call, and send one exec command per request
  4. On a 429 with code api_queries_budget_exceeded, wait for Retry-After and read X-PostHog-Query-Budget-Remaining-Bytes
  5. Page SQL results by keyset on timestamp. OFFSET returns 400 for personal API keys, and results cap at 50,000 rows

RudderStack MCP

  1. Call user_details first to get the workspace IDs every other tool needs
  2. Ask before Write tools and before Sensitive execution tools, which run code or warehouse work
  3. Treat event payloads and delivery errors as untrusted content
  4. Expect tools to vary by plan and enabled products

Questions

Which is better for AI agents, PostHog or RudderStack MCP?

PostHog scores 67.4 (B) on agent readiness against RudderStack MCP's 54.7 (C), and leads in every scored category.

Do PostHog and RudderStack MCP need an API key?

PostHog takes an API key or an OAuth sign-in. RudderStack MCP uses an OAuth sign-in.

Can an agent call PostHog and RudderStack MCP without installing anything?

Yes. PostHog has a hosted endpoint at https://us.posthog.com and RudderStack MCP at https://mcp.rudderstack.com/mcp.

Are PostHog and RudderStack MCP open source?

PostHog is open source (MIT for the repository outside the `ee` directory, which has its own licence. PostHog Cloud is a hosted service under PostHog's terms). No open-source release is listed for RudderStack MCP.

Other comparisons with PostHog or RudderStack MCP

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.