Hello, We operate the Markovian Protocol witness, listed in the testing table since July 29, so about six weeks (about page: https://witness.markovianprotocol.com/about). We'd like to be considered for the staging tier. We would configure testing/log-list.1 plus both staging lists (log-list-10qps-4klogs.1 and log-list-100qps-40klogs.1), refresh hourly, and keep the never-remove semantics we already follow. Capacity, measured August 2 on an isolated instance of the same witness build on the same host, loaded with a 40,000-entry known-logs table: 10, 25, 50, and 100 requests per second sustained for 45 seconds each, zero errors across 12,800 valid signed checkpoints with real consistency proofs; median latency 5-6 ms and p99 under 10 ms at every sustained rate; an unthrottled burst averaged 315 requests per second, also error-free. The known-logs table size had no measurable effect on latency. Operations, stated plainly: single operator, single host (Hetzner, Helsinki), SQLite state with an off-host backup, Cloudflare fronting the public endpoint, C2SP refusal rules (404/403/409/422) enforced. The cloudflared process fronting the endpoint has been up since 2026-08-31, 8 days, and reports 222,129 proxied requests with zero request errors. That counter resets when the process restarts, so it is 8 days of evidence, not six weeks -- we are not claiming an unbroken record behind it. If there's anything else you'd like to see first, tell us and we'll do it. Markovian Protocol hello@markovianprotocol.com