Block a user
Fix actual_tier never updated from "unknown" in run_agent_task
Benchmark: complex tier never triggered — 0% accuracy (40 queries)
Benchmark: smart home commands (medium) mis-routed to light
Benchmark: light tier over-classified as medium (tech definition queries)
Fix routing: add Russian tech def patterns to light, strengthen medium smart home
Remove or replace Bifrost test in test_memory.py
Remove Bifrost: replace test 4 with LiteLLM health check
Fix tier logging: capture actual_tier, fix parse_run_block regex, remove reply_text truncation
Fix benchmark log extraction: first tier match, increase log tail to 300
Fix tier logging: capture actual_tier, fix parse_run_block regex, remove reply_text truncation
Fix benchmark log extraction: first tier match, increase log tail to 300