alvis/oO

M2 — AI tips + multi-source signals

New Issue

Goal: tips are AI-generated from user context, not just raw Todoist tasks. Multiple signal sources feed a generalized pipeline. Research-intensive milestone.

Key themes:

Signal source abstraction (beyond Todoist)
AI tip generation via local LLM (Ollama)
Generalized candidate → rank → render pipeline
Feature registry + profile builder
Policy research (Thompson sampling, neural bandits)
Reward model improvements

No due date

76% Completed

finish the mechanics of active notifications of the forgotten tasks

#126 opened 2026-05-11 10:23:53 +00:00 by alvis

need to add openclaw and hermes memory systesm

#124 opened 2026-05-11 10:08:22 +00:00 by alvis

integrate google health

#119 opened 2026-05-11 07:48:28 +00:00 by alvis

random policy should return random tasks, not todoist tasks

#117 opened 2026-05-06 10:24:07 +00:00 by alvis

meta: shared context-inference framework for agents

#111 opened 2026-05-05 09:24:29 +00:00 by alvis

Advice library + mixing policy ml

#32 opened 2026-04-13 14:23:20 +00:00 by alvis

Client-side timing signals: location, idle, focus windows ml web

#31 opened 2026-04-13 14:23:20 +00:00 by alvis

profile: unified Profile model (prefs, contexts, consents) backend

#30 opened 2026-04-13 14:23:20 +00:00 by alvis

integrations: generic webhook ingress backend

#29 opened 2026-04-13 14:23:20 +00:00 by alvis

integrations: Apple Health (web import) connector backend

#28 opened 2026-04-13 14:23:20 +00:00 by alvis

integrations: Google Calendar connector backend

#27 opened 2026-04-13 14:23:20 +00:00 by alvis

Phase 0 E2E: Playwright suite covering sign-in → connect Todoist → see a tip platform

#20 opened 2026-04-13 14:23:20 +00:00 by alvis

auth: Apple OAuth sign-in backend

#7 opened 2026-04-13 14:23:19 +00:00 by alvis

Gitea Actions CI skeleton infra platform

#3 opened 2026-04-13 14:23:19 +00:00 by alvis