Busabase
Busabase is an open-source (MIT) database and workspace for AI agents — data, docs, skills, and apps in one place. Run it locally, self-host it inside your network, or use Busabase Cloud, and turn age
翻譯準備中
尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權,自行管理介紹,無需管理員核准。
登入並產生驗證檔案,在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案,即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入,將前往登入頁面。登入後請回到此處,再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify,即可驗證所有權並自行管理介紹。
一分鐘完成驗證Independent civic hazard intelligence platform for Hawaiʻi — real-time wildfire, smoke, flood, air quality and volcanic situational awareness from NASA FIRMS, NOAA, NWS, EPA and PacIOOS.
翻譯準備中
正在載入完整 README。 以下為已儲存的摘錄。
Production Rule: All live Kahu Ola surfaces must use real upstream hazard data only. Mock states are allowed only for development review and must never be shipped to production. Kahu Ola Hawaiʻi Civic Hazard Intelligence Platform Kahu Ola is a civic hazard intelligence platform for Hawaiʻi designed to transform complex public hazard data into clear situational awareness signals for residents, responders, and community organizations. The system aggregates data from multiple public sources including:檢視完整 README ↗
Production Rule: All live Kahu Ola surfaces must use real upstream hazard data only. Mock states are allowed only for development review and must never be shipped to production.
Kahu Ola Hawaiʻi Civic Hazard Intelligence Platform
Kahu Ola is a civic hazard intelligence platform for Hawaiʻi designed to transform complex public hazard data into clear situational awareness signals for residents, responders, and community organizations.
The system aggregates data from multiple public sources including:
NASA FIRMS wildfire satellite detections
National Weather Service alerts
NOAA radar and storm data
EPA AirNow air quality feeds
USGS Hawaiian Volcano Observatory
PacIOOS ocean and coastal sensors
RAWS and MesoWest weather stations
Kahu Ola converts these inputs into validated hazard signals presented through a calm, civic-grade interface designed for rapid comprehension during emergencies.
The platform does not replace official emergency systems and never issues evacuation instructions.
Instead, it provides situational awareness signals and directs users to official emergency authorities.
System Philosophy
Kahu Ola follows a simple principle:
Complex hazard data should be understandable in five seconds.
The system prioritizes:
plain-language hazard signals
freshness transparency
deterministic hazard priority
civic-safe action links
progressive disclosure of technical sources
High-Level Architecture External Hazard Sources (NASA · NOAA · NWS · EPA · USGS)
↓
Pollers / Parsers (signal acquisition)
↓
Validation Layer (fail-closed parsing)
↓
Signal Normalization (FireSignal, FloodSignal, etc.)
↓
Freshness Engine (FRESH / STALE_OK / STALE_DROP)
↓
Event Priority Engine (multi-hazard decision)
↓
Snapshot Cache (redis / edge cache)
↓
Cloudflare Worker API
↓
Clients Web · Mobile · Smartwatch · Widgets
The client layer renders server-validated hazard signals.
Clients never interpret raw hazard data directly.
Repository Structure kahuola/ │ ├── docs/ # system specifications │ │ ├── doctrine/ │ ├── architecture/ │ ├── api/ │ ├── map/ │ ├── ux/ │ └── runtime/ │ ├── worker/ # Cloudflare Worker aggregator │ ├── web/ # website frontend │ ├── mobile/ # mobile client applications │ ├── scripts/ # pollers and snapshot builders │ └── infrastructure/ # deployment configuration Documentation Index
All system specifications live under /docs.
New contributors should begin reading in this order.
Primary system design and hazard data pipeline.
docs/architecture/V4_8_MASTER_INDEX.md docs/architecture/V4_8_SYSTEM_ARCHITECTURE.md docs/architecture/V4_8_HAZARD_SIGNAL_PIPELINE.md docs/architecture/V4_8_CLIENT_ARCHITECTURE.md
These documents describe:
overall system architecture
hazard signal pipeline
client architecture model
Defines how hazards are interpreted and prioritized.
docs/runtime/V4_8_STATE_MACHINE.md docs/runtime/V4_8_EVENT_PRIORITY_SPEC.md docs/runtime/V4_8_MULTI_HAZARD_DECISION_TREE.md docs/runtime/V4_8_MOCK_STATE_SPEC.md
These documents define:
hero signal selection
multi-hazard conflict resolution
degraded system behavior
Defines data structures shared by all clients.
docs/api/V4_8_API_CONTRACT.md docs/api/V4_8_SIGNAL_MODEL_SPEC.md docs/api/V4_8_DATA_FRESHNESS_POLICY.md
These specifications define:
stable API endpoints
unified signal models
freshness classification
Defines how hazard information is presented.
docs/ux/V4_8_HOMEPAGE_HERO_SPEC.md docs/ux/V4_8_ALERT_COPY_GUIDELINES.md docs/ux/V4_8_DEGRADED_STATE_UX_SPEC.md
These documents define:
hero signal design
alert copy standards
degraded state user experience
Defines hazard map layers and rendering order.
docs/map/V4_8_MAP_LAYER_SPEC.md
Includes:
wildfire hotspots
rainfall radar
flood polygons
smoke plume overlays
Defines server aggregation and scaling model.
docs/architecture/V4_8_WORKER_AGGREGATOR_SPEC.md docs/architecture/V4_8_CACHE_STRATEGY.md docs/architecture/V4_8_RATE_LIMIT_PROTECTION.md
These documents define:
Cloudflare Worker aggregation
caching strategy
rate-limit protection
Defines non-negotiable architectural principles.
docs/doctrine/Kahu_Ola_V4_7_Doctrine.md docs/doctrine/V4_8_RUNTIME_PROTECTION_ADDENDUM.md
Key invariants include:
client stateless architecture
fail-closed data validation
civic trust guarantees
degraded state transparency
Operational Guarantees
Kahu Ola is designed to remain trustworthy even when upstream systems fail.
The platform provides the following guarantees.
Client Stateless Guarantee
Clients never query upstream hazard providers directly.
Client → Worker → Upstream APIs
All hazard interpretation occurs server-side.
Cache-First Hazard Delivery
Hazard signals are served from validated cached snapshots, not live upstream queries.
This prevents upstream outages from breaking the interface.
Freshness Transparency
Every signal contains explicit freshness metadata.
States include:
FRESH STALE_OK STALE_DROP
Users are never shown stale data without labeling.
Fail-Closed Parsing
Invalid upstream payloads are dropped entirely.
The system never attempts to repair malformed hazard data.
Degraded Mode
When upstream sources fail, the platform enters degraded mode.
Example message:
Some hazard data sources are delayed. Showing the last verified snapshot where possible.
The system never silently hides degraded conditions.
Deterministic Hazard Priority
Multiple hazards are resolved using a deterministic priority model.
Example order:
Degraded Fire Active Flood Warning Flood Watch Storm Active Volcanic Watch Monitoring
Full specification:
docs/runtime/V4_8_MULTI_HAZARD_DECISION_TREE.md Civic Authority Boundary
Kahu Ola does not issue emergency orders.
The system never:
instructs evacuation routes
overrides official alerts
predicts disaster outcomes
Users must follow guidance from:
Hawaiʻi Emergency Management Agency
County emergency management offices
National Weather Service
other official authorities
Hazard Data Sources
Kahu Ola integrates public hazard data from multiple agencies.
Primary sources include:
NASA FIRMS wildfire hotspots
NOAA radar and weather data
National Weather Service alerts
EPA AirNow air quality
NOAA HMS smoke plume detection
NIFC fire perimeters
PacIOOS coastal sensors
RAWS fire weather stations
USGS Hawaiian Volcano Observatory
These sources are normalized through the Kahu Ola signal pipeline before being presented to users.
System Goal
Kahu Ola aims to provide clear hazard awareness without replacing official emergency systems.
The platform focuses on:
clarity
transparency
civic trust
resilience during disasters
By translating complex hazard data into simple signals, Kahu Ola helps Hawaiʻi residents understand evolving hazard conditions quickly and safely.
License
See the LICENSE file for project license information.
Contributing
Contributors should review the architecture documentation before submitting changes.
Start with:
docs/architecture/V4_8_MASTER_INDEX.md
All system behavior must follow the V4 doctrine and associated runtime specifications.
All production Kahu Ola pages must use real upstream hazard data only. Mock data, placeholder endpoints, and demo states are forbidden in production. Browser clients must never call upstream hazard APIs directly. All signals must pass through validation, freshness classification, and cache before rendering. Stale or degraded data must always be labeled. No future contributor may reintroduce known V4-fixed bugs:
Kahu Ola is a civic hazard intelligence platform for Hawaiʻi designed to provide clear, real-time situational awareness during natural hazards such as wildfires, floods, storms, volcanic activity, and air quality events.
The system translates complex satellite and meteorological data into plain-language hazard signals that residents can understand within seconds.
Primary mission:
E mālama pono — protect life through clear information.
Kahu Ola follows a strict architecture doctrine to ensure reliability, transparency, and public trust.
All hazard data must flow through the Kahu Ola Aggregator.
Browser / mobile apps must never call upstream hazard APIs directly.
Client → Cloudflare Worker → Upstream Sources
Upstream sources include:
Production systems must only use real hazard data.
Mock data is strictly forbidden in production environments.
Mock states are allowed only for development previews.
The system must never show a blank screen.
If upstream data fails:
FRESH → STALE_OK → STALE_DROP → DEGRADED
Users must always see the current system state.
Kahu Ola provides situational awareness only.
It does not issue evacuation orders.
Official emergency instructions always come from:
All system documentation is organized in:
docs/
See:
docs/architecture/V4_8_MASTER_INDEX.md
See LICENSE file.
See CONTRIBUTING.md
Start here:
docs/PROJECT_PRIMER.md docs/AI_ONBOARDING.md docs/architecture/V4_8_MASTER_INDEX.md