Findings · measured, dated, recomputable

78.7% of observed MCP tool-surface changes carried no version change.

How often does an MCP server change the tools it exposes without changing its version number?

Cite as
78.7% of MCP tool-surface changes carry no version change (Still OS Digital Holdings LLC, Surface History, n=6,384 changes across 4,680 hosts, 2026-09-21 to 2026-10-30).

The numbers

Surface changes observed6,384
Of those, no version change5,025 (78.7%)
Distinct hosts observed changing4,680
Servers in the observed universe22,295
Observation window2026-09-21 → 2026-10-30

Why it matters

Version pinning is the standard control for dependency risk. These observations indicate the version string does not signal a change in what an MCP server can actually do, so a pinned version is not evidence of a fixed tool surface.

Method

Each server's live tool list is fetched on a repeating cadence and reduced to SHA256(JSON.stringify(sorted tool names)). A change in that digest with no change in the reported server version is counted as version-silent. Event-sourced: any past date is replayable from the event log.

Producer

The files these figures were derived from, on the machine that published this page

/home/marcus/state/mcp-census/surface-clock/surface-events.jsonl
/home/marcus/state/mcp-census/surface-clock/latest-surfaces.json

Next

Check one host's record

Built 2026-10-30T22:17:03.185Z · JSON