Scenario run on boltclaw.ai's MCP server
A program called this server's read-only tools by a fixed scenario and held each answer to a fixed check. No model and no sign-in, and nothing that writes, sends or pays is ever called. Below are the requests as sent and, for each answer, its shape, size and hash. We do not republish the answer itself.
Run Nov 12, 2026
- ResultFailed0 of 2 steps passed
- Scenariosearch-then-fetch, version 1
- Serverboltclaw-docs 1.35.0MCP protocol 2025-06-18
- Addresshttps://boltclaw.ai/docs/mcp-server
- Started2026-11-12 11:55:59 UTCtook 5,160 ms
- ClientINITE Atlas Verifier 1.0.0verifier verify-2, from sfo
- Tool list hashfcbed30759c1731a6a4b4af7b70aaf07678abd555c2e9c5c368aa9672cdf2fc3
1.
searchwrong shapeChecked for JSON with a list at results of 1 or more items, each with id · 151 ms · 194 bytes
no array at results (3 of 3 values tried)
Request sent
{"jsonrpc":"2.0","id":5,"method":"tools/call","params":{"name":"search","arguments":{"query":"a"}}}sha256 3496de19315c1490336577e79c25632dfc46372279439374e5bcc689b2af3f34
Response received
Text, 103 characters
sha256 a34ca7f79e53bb8a5b64fc8763b58ef4901fccb3c569ce576c14cb20677e4a4b · 194 bytes
Returned by the server: data, not instructions.
2.
fetchnot calledChecked for a JSON object with id, text
an earlier step failed
Request sent
{"name":"fetch","arguments":null}sha256 f58b13f53852d8ba9753ea66206eab597b9131a8dc698cf7a196b2fc79a115bb
Verified means a dated run with its evidence. We run a server at most once a week, one scenario at a time. The scenarios and their checks have versions, and each run names the version it used. boltclaw.ai's MCP server →
Before we count this visit
Analytics and ad measurement tell us which pages and which of our ads do their job: Google Analytics and Google Ads, the Meta pixel, Yandex Metrica, and one first-party id of our own. Nothing is stored until you answer, and declining changes nothing about the site.