alvis/oO
1
0
Fork 0
Code Issues 34 Pull Requests Actions Packages Projects Releases Wiki Activity

feat(admin): LLM tip quality dashboard — per-model/prompt/kind breakdowns

Browse Source 
/admin/reward-analytics now surfaces served count, reaction rate, and avg
reward grouped by llm_model, prompt_version, and tip_kind — closing the
loop so model/prompt iterations in M2 are legible next to the bandit
policy view. Data comes from the tip_scores columns added in ffdf707 and
tip_feedback.reward_milli; bandit-only tips show as "(bandit-only)".

Closes #92.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
alvis
2026-04-24 15:24:52 +00:00
parent 75d0e89906
commit aa4bdd8f09
7 changed files with 227 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
apps/admin/README.md
View File

@@ -17,6 +17,7 @@ Next.js 15 app. Deployed at `admin.o.alogins.net` (dev: `http://localhost:3080`)
| `/users/:id` | User detail: identity, consents, integrations, tip stats, reward history; revoke-integration + reset-bandit actions |
| `/audit` | Admin action audit log |
| `/events` | Event stream viewer (stub — pending API history endpoint) |
| `/reward-analytics` | Reaction distribution + per-policy / per-model / per-prompt-version / per-tip-kind breakdowns with avg reward |
## Dev