tinyhumansai/openhuman v0.50.3

OpenHuman v0.50.3

on GitHub

What's Changed

• refactor: replace dynamic imports with static imports in app/src by @graycyrus in #111
• feat(web-search): add Parallel search provider to WebSearchTool by @YellowSnnowmann in #114
• Feat/memory agent by @graycyrus in #115
• refactor(config): centralize env vars and add .env.example docs by @graycyrus in #116
• Add core memory RPC surface (#69) by @sanil-23 in #112
• feat: add Discord notification for new releases in GitHub Actions wor… by @YellowSnnowmann in #118
• feat(local-ai): guided model tier selection by device capability by @graycyrus in #120
• update: modify .gitignore to include create_issue and remove redund… by @graycyrus in #121
• Add GLiNER relex ingestion pipeline (#60) by @sanil-23 in #113
• feat(installer): add one-command installer flow and release smoke checks by @oxoxDev in #123
• feat(autocomplete): persist accepted completions in memory and reuse them for suggestions (#108) by @oxoxDev in #119
• fix(agent): execute fallback tool calls in the loop by @M3gA-Mind in #125
• feat: re-setup skills runtime and add model-callable skills bridge by @M3gA-Mind in #126
• fix(tauri): restore working tray controls in desktop shell by @M3gA-Mind in #127
• feat(memory): connect graph query and doc ingest APIs by @sanil-23 in #124
• test: add cross-stack coverage for core and tauri flows by @M3gA-Mind in #130
• feat(observability): add Sentry error reporting to Rust core by @senamakel in #131
• fix: propagate entity_type into retrieval context for graph visualization by @sanil-23 in #135
• fix(e2e): harden deep-link login flow reliability by @M3gA-Mind in #137
• fix(skills): enforce per-skill runtime tool isolation by @M3gA-Mind in #140
• fix: service gate buttons unclickable on startup by @senamakel in #139
• refactor: migrate memory service to controller registry by @sanil-23 in #138
• feat(onboarding): calmer onboarding + local AI download snackbar by @graycyrus in #134
• ci: speed up GitHub Actions builds (~14m → ~3-5m warm) by @senamakel in #136
• fix(local-ai): Ollama bootstrap failure UX and auto-recovery by @senamakel in #142
• fix(skills): persist OAuth credentials and fix skill auto-start lifecycle by @senamakel in #146
• Update issue templates by @senamakel in #148
• feat(agent): add self-learning subsystem with post-turn reflection by @senamakel in #149
• feat(auth): Telegram bot registration flow — /auth/telegram endpoint by @senamakel in #150
• feat(webhooks): webhook tunnel routing for skills + remove legacy tunnel module by @senamakel in #147
• feat(agent): architecture improvements — context guard, cost tracking, permissions, events by @senamakel in #151
• refactor(models): standardize to reasoning-v1, agentic-v1, coding-v1 by @senamakel in #152
• fix(skills): debug infrastructure + disconnect credential cleanup by @senamakel in #154
• feat(agent): multi-agent harness with 8 archetypes, DAG planning, and episodic memory by @senamakel in #155

New Contributors

• @YellowSnnowmann made their first contribution in #114
• @sanil-23 made their first contribution in #112

Full Changelog: v0.49.32...v0.50.3