llm-team-ui

Go to file

root 28df789745 Fix runaway experiments: cap at 50 trials, fix DB permissions

Bugs fixed:
- Ratchet loop had no trial cap — experiment #1 ran 3762 trials
  unchecked. Now capped at max_trials=50 per start cycle.
- meta_pipelines, meta_runs, self_reports tables had no GRANT
  for kbuser — fixed permissions for all tables and sequences.

All 4 running experiments auto-paused on restart.
Stress test confirms all tables accessible, all models responding,
meta-pipeline creation working, self-report save/retrieve working.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-26 04:56:37 -05:00

server

Add triage, backup, and disaster recovery system

2026-03-25 04:52:48 -05:00

.gitignore

LLM Team UI v1.0 — full-stack local AI orchestration platform

2026-03-25 02:51:36 -05:00

llm_team_config.json

LLM Team UI v1.0 — full-stack local AI orchestration platform

2026-03-25 02:51:36 -05:00

llm_team_ui.py

Fix runaway experiments: cap at 50 trials, fix DB permissions