Field-by-field comparison
Compare the evidence before choosing an MCP.
Each signal stays separate, missing facts remain visible, and the column order follows your selection.
| Evidence field | Statsig MCPShortlist candidate | Salesforce Hosted MCPMCP100 Recommended |
|---|---|---|
| Task fit | ||
| Best for |
|
|
| Not ideal for |
|
|
| Avoid when |
|
|
| Provenance | ||
| Provenance details | First-party MCP Statsig · Publisher source ↗ | First-party MCP Salesforce · Publisher source ↗ |
| Maintenance | ||
| Maintenance details | Repo: Not documented Package: Not documented | Repo: Not documented Package: Not documented |
| Popularity evidence | ||
| Client coverage | ||
| Client coverage details | Not documented | Claude Code✓ Streamable HTTP Claude Desktop✓ Streamable HTTP Codex✓ Streamable HTTP VS Code✓ Streamable HTTP Cursor✓ Streamable HTTP OpenCode✓ Streamable HTTP |
| Client coverage note | Local stdio runs on your computer. Streamable HTTP connects to a remotely hosted MCP server. | |
| Access and data | ||
| Authentication | OAuth for the hosted V3 MCP; a Console API key is also supported in documented setup paths. Write operations require credentials with write permission. | Follow only the publisher source; credential storage, scopes, and revocation remain under review. |
| Cost | No separate MCP fee is documented. Statsig has a free Developer plan and paid Pro/Enterprise plans with event-volume limits and overage/custom pricing. | Publisher plans, API usage, infrastructure, or per-action charges may apply; MCP100 has not published a cost conclusion. |
| Permissions | Tool visibility and executable operations depend on Statsig project permissions and server configuration. Project permissions and review policies continue to apply. | Exact permissions and read/write boundaries have not completed MCP100 review. |
| Data handling | Not documented in the reviewed sources. The reviewed Statsig MCP documentation does not state an MCP-specific prompt or tool-result retention/training policy. | Candidate-specific retention, subprocess, telemetry, and model-data behavior remain review items. |
| Limitations | ||
| Tradeoffs |
|
|
| Risk context | V3 includes read, create, update and destructive/review operation lanes. Changes can affect feature gates, experiments, configs and production rollout behavior. | Use a test account, least privilege, read-only controls where available, and explicit approval for mutations. |
| Evidence date | ||
| Editorial review | 2026-10-19 | 2026-07-24 |
| Candidate evidence | 2026-10-20 | 2026-09-24T04:25:05.554391Z |
Popularity, maintenance, fit, permissions, and client support are independent evidence fields.
Read the evidence method →