Red Tide Algal Bloom Monitoring & Early Warning Skill
Skill Overview
This skill transforms Claude into a domain-expert for Harmful Algal Bloom (HAB) Monitoring & Aquaculture Safety. It delivers structured, evidence-backed outputs by combining real-time data aggregation, recognized domain methods, and academic research into a single orchestrated workflow with explicit risk disclosure.
Core Capabilities
- Bloom Detection: Satellite (Sentinel-3 OLCI, MODIS) + in-situ sensor analysis
- Species Identification: Karenia, Alexandrium, Dinophysis, Pseudo-nitzschia, cyanobacteria
- Toxin Risk Assessment: Paralytic, diarrhetic, amnesic, neurotoxic shellfish poisoning
- Aquaculture Impact: Shellfish vulnerability, harvest closure timing, economic risk
- Environmental Triggers: Temperature, nutrients (N/P/Si), stratification, upwelling
- Early Warning: Alert levels, response protocols, mitigation strategies
Evidence-Based Framework
All outputs cite ≥3 sources with ≥1 academic/authoritative reference, explicit evidence hierarchy (Tier 1-4), and mandatory risk/limitation disclosure. The skill continuously improves through automated knowledge crawling from academic databases and domain authorities.
Harness Architecture
Skill Registry Pattern
The skill implements a flexible agent & skill architecture using a modular pattern:
/skill activation
│
├─ Language Detection (pre-flight)
│ ├─ Vietnamese → translation + bilingual output
│ ├─ English → direct processing
│ └─ Other → English processing + translation offer
│
├─ Step 1: sub-gather-requirements
│ └─ Input: User query
│ └─ Output: Structured requirements (object, scope, timeframe, inputs, audience, language)
│
├─ Step 2: sub-evidence-collector
│ ├─ Fetch: Current satellite data (Copernicus, NOAA)
│ ├─ Fetch: In-situ sensor data (fluorometer, CTD)
│ ├─ Fetch: Authoritative documents (IOC-UNESCO, WOAH)
│ └─ Output: Evidence bundle with sources + dates
│
├─ Step 3: sub-core-analysis
│ ├─ Detect: Satellite chl-a + SST analysis
│ ├─ Detect: In-situ cell counts + species ID
│ ├─ Analyze: Environmental triggers (T, nutrients, stratification)
│ ├─ Assess: Toxin risk (type, concentration, regulatory limits)
│ └─ Alert: Aquaculture impact + response levels
│
├─ Step 4: sub-knowledge-updater
│ ├─ Query: SECOND-KNOWLEDGE-BRAIN.md
│ ├─ Surface: Academic citations (with Tier labels)
│ └─ Flag: Knowledge gaps for crawl pipeline
│
├─ Step 5: sub-advisor
│ ├─ Synthesize: All prior analysis
│ ├─ Conclude: Risk category (Low/Monitor/Critical)
│ ├─ Recommend: Specific actions with evidence chain
│ └─ Disclose: Limitations + uncertainty
│
└─ Step 6: main quality gate
├─ Verify: Evidence hierarchy compliance
├─ Verify: Source citations (≥3 sources, ≥1 authoritative)
├─ Verify: Risk disclosure present
├─ Verify: Output template completeness
└─ Auto-fix: Up to 2 retries for missing elements
Agent Registration & Resolution
Skill Registration
Skills are registered in /skills/*.md with mandatory frontmatter:
---
name: sub-gather-requirements
description: Clarify the object of analysis, constraints, timeframe, available inputs, target audience, and language before any data fetching.
---
Registration Requirements:
name: Unique identifier (kebab-case)description: One-line summary (mandatory, used for routing)- Sections: Role & Persona, Workflow, Tools, Output Format, Quality Gates
Skill Resolution
The harness uses semantic routing to resolve the appropriate sub-skill:
- Input Analysis: Extract task type from user query
- Skill Matching: Compare against skill descriptions
- Execution Order: Follow sequential harness flow
- Fallback Chain: On failure → knowledge base → simplified analysis → explicit limitation flag
Skill Execution Model
Each sub-skill follows the execution lifecycle:
class SubSkillExecution:
def execute(self, context: ExecutionContext) -> SkillResult:
# 1. Input validation
self._validate_inputs(context)
# 2. Tool execution (with retries)
result = self._execute_tools(context)
# 3. Quality gate validation
self._validate_quality_gates(result)
# 4. Error handling (graceful degradation)
if result.failed:
return self._handle_degradation(context)
# 5. Output formatting
return self._format_output(result)
Input/Output JSON Schemas
Input Schema (User Query)
{
"$schema": "http://json-schema.org/draft-07/schema#",
"title": "RedTideSkillInput",
"type": "object",
"required": ["query"],
"properties": {
"query": {
"type": "string",
"description": "User's natural language query about HAB monitoring or aquaculture safety",
"minLength": 1,
"examples": [
"Analyze red tide risk for oyster farms in Apalachicola Bay",
"What's the current bloom status off the coast of Maine?",
"Assess toxin risk for shellfish harvest in Puget Sound"
]
},
"context": {
"type": "object",
"properties": {
"region": {"type": "string", "description": "Geographic region of interest"},
"timeframe": {"type": "string", "description": "Analysis timeframe (e.g., 'current', 'next 7 days')"},
"target_audience": {"type": "string", "enum": ["aquaculture", "public_health", "researcher", "general"]},
"language": {"type": "string", "enum": ["en", "vi", "es", "fr", "ja", "zh"], "default": "en"}
}
},
"options": {
"type": "object",
"properties": {
"include_recommendations": {"type": "boolean", "default": true},
"include_scenarios": {"type": "boolean", "default": true},
"detail_level": {"type": "string", "enum": ["brief", "standard", "comprehensive"], "default": "standard"}
}
}
}
}
Output Schema (Skill Result)
{
"$schema": "http://json-schema.org/draft-07/schema#",
"title": "RedTideSkillOutput",
"type": "object",
"required": ["metadata", "analysis", "evidence_chain", "conclusion", "disclosure"],
"properties": {
"metadata": {
"type": "object",
"required": ["skill_version", "timestamp", "language", "processing_time_ms"],
"properties": {
"skill_version": {"type": "string", "pattern": "^\\d+\\.\\d+\\.\\d+$"},
"timestamp": {"type": "string", "format": "date-time"},
"language": {"type": "string"},
"processing_time_ms": {"type": "number"},
"degradation_level": {"type": "integer", "minimum": 0, "maximum": 4}
}
},
"requirements": {
"type": "object",
"properties": {
"object_of_analysis": {"type": "string"},
"scope": {"type": "string"},
"timeframe": {"type": "string"},
"available_inputs": {"type": "array", "items": {"type": "string"}},
"target_audience": {"type": "string"},
"language": {"type": "string"},
"analysis_type": {"type": "string"}
}
},
"evidence_bundle": {
"type": "object",
"properties": {
"current_data": {
"type": "object",
"properties": {
"satellite_chl_a": {"type": "object"},
"sst": {"type": "object"},
"in_situ_counts": {"type": "object"},
"toxin_levels": {"type": "object"}
}
},
"authoritative_docs": {
"type": "array",
"items": {
"type": "object",
"properties": {
"source": {"type": "string"},
"url": {"type": "string"},
"date": {"type": "string"},
"tier": {"type": "integer", "minimum": 1, "maximum": 4}
}
}
},
"recent_news": {"type": "array", "items": {"type": "object"}},
"reference_benchmarks": {"type": "object"}
}
},
"analysis": {
"type": "object",
"properties": {
"detection": {
"type": "object",
"properties": {
"bloom_status": {"type": "string"},
"cell_concentration": {"type": "string"},
"species_identified": {"type": "array", "items": {"type": "string"}},
"confidence": {"type": "string"}
}
},
"triggers": {
"type": "object",
"properties": {
"temperature": {"type": "object"},
"nutrients": {"type": "object"},
"stratification": {"type": "string"},
"upwelling": {"type": "string"}
}
},
"toxin_risk": {
"type": "object",
"properties": {
"toxin_type": {"type": "array", "items": {"type": "string"}},
"risk_level": {"type": "string"},
"regulatory_limit_exceeded": {"type": "boolean"}
}
},
"aquaculture_impact": {
"type": "object",
"properties": {
"species_affected": {"type": "array", "items": {"type": "string"}},
"vulnerability_level": {"type": "string"},
"economic_risk": {"type": "string"}
}
}
}
},
"knowledge_base": {
"type": "array",
"items": {
"type": "object",
"properties": {
"citation": {"type": "string"},
"tier": {"type": "integer", "minimum": 1, "maximum": 4},
"relevance": {"type": "string"}
}
}
},
"evidence_chain": {
"type": "array",
"items": {
"type": "object",
"properties": {
"claim": {"type": "string"},
"evidence": {"type": "array", "items": {"type": "string"}},
"confidence": {"type": "string"}
}
}
},
"conclusion": {
"type": "object",
"required": ["category"],
"properties": {
"category": {
"type": "string",
"enum": [
"Low Risk",
"Normal",
"Monitor (elevated)",
"Critical Bloom Alert",
"Inconclusive"
]
},
"summary": {"type": "string"},
"confidence": {"type": "string"},
"key_factors": {"type": "array", "items": {"type": "string"}}
}
},
"recommendations": {
"type": "array",
"items": {
"type": "object",
"properties": {
"action": {"type": "string"},
"priority": {"type": "string"},
"evidence": {"type": "array", "items": {"type": "string"}}
}
}
},
"scenarios": {
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {"type": "string"},
"probability": {"type": "string"},
"impact": {"type": "string"},
"mitigation": {"type": "array", "items": {"type": "string"}}
}
}
},
"disclosure": {
"type": "object",
"required": ["limitations", "uncertainties", "data_quality"],
"properties": {
"limitations": {"type": "array", "items": {"type": "string"}},
"uncertainties": {"type": "array", "items": {"type": "string"}},
"data_quality": {"type": "string"},
"assumptions": {"type": "array", "items": {"type": "string"}}
}
}
}
}
Tool Definitions & Schemas
Tool Registry
The skill defines rich tool definitions with schemas and execution handlers:
Satellite Data Fetch Tool
{
"tool_name": "fetch_satellite_data",
"description": "Fetch current satellite chlorophyll-a and sea surface temperature data for specified region",
"parameters": {
"region": {
"type": "string",
"description": "Geographic region (bounding box or named region)",
"required": true
},
"date_range": {
"type": "string",
"description": "Date range for data retrieval (ISO 8601 format)",
"required": false,
"default": "last_7_days"
},
"resolution": {
"type": "string",
"enum": ["1km", "4km", "9km"],
"default": "4km"
}
},
"response_schema": {
"chl_a": {
"mean": "float",
"max": "float",
"anomaly": "float",
"data_url": "string"
},
"sst": {
"mean": "float",
"min": "float",
"max": "float",
"anomaly": "float",
"data_url": "string"
},
"metadata": {
"satellite": "string",
"processing_date": "string",
"coverage": "float"
}
},
"execution_handler": "sub-evidence-collector",
"timeout": 30,
"retry": 3
}
In-Situ Sensor Query Tool
{
"tool_name": "query_in_situ_sensors",
"description": "Query in-situ monitoring stations for cell counts and environmental parameters",
"parameters": {
"region": {"type": "string", "required": true},
"sensor_types": {
"type": "array",
"items": {"enum": ["fluorometer", "ctd", "spectrometer"]},
"default": ["fluorometer"]
},
"timeframe": {"type": "string", "default": "last_7_days"}
},
"response_schema": {
"stations": [
{
"station_id": "string",
"location": {"lat": "float", "lon": "float"},
"measurements": {
"cell_count": {"value": "int", "units": "cells/L"},
"chlorophyll": {"value": "float", "units": "mg/m³"},
"temperature": {"value": "float", "units": "°C"},
"salinity": {"value": "float", "units": "PSU"}
},
"timestamp": "string"
}
],
"metadata": {
"station_count": "int",
"data_quality": "string"
}
},
"execution_handler": "sub-evidence-collector",
"timeout": 20,
"retry": 2
}
Authoritative Document Fetch Tool
{
"tool_name": "fetch_authoritative_docs",
"description": "Fetch current guidance and standards from authoritative HAB monitoring sources",
"parameters": {
"sources": {
"type": "array",
"items": {"enum": ["IOC-UNESCO", "NOAA", "Copernicus", "WOAH"]},
"default": ["IOC-UNESCO", "NOAA"]
},
"topic": {"type": "string", "required": true},
"document_types": {
"type": "array",
"items": {"enum": ["guidelines", "standards", "reports", "alerts"]},
"default": ["guidelines", "alerts"]
}
},
"response_schema": {
"documents": [
{
"source": "string",
"title": "string",
"url": "string",
"publication_date": "string",
"tier": "int",
"summary": "string",
"key_points": ["string"]
}
],
"metadata": {
"document_count": "int",
"source_coverage": "string"
}
},
"execution_handler": "sub-evidence-collector",
"timeout": 45,
"retry": 3
}
Knowledge Base Query Tool
{
"tool_name": "query_knowledge_base",
"description": "Query SECOND-KNOWLEDGE-BRAIN.md for academic and professional evidence",
"parameters": {
"keywords": {
"type": "array",
"items": {"type": "string"},
"required": true
},
"max_results": {"type": "integer", "default": 5, "maximum": 10},
"min_tier": {"type": "integer", "enum": [1, 2, 3, 4], "default": 2}
},
"response_schema": {
"results": [
{
"citation": "string",
"tier": "int",
"relevance_score": "float",
"year": "int",
"venue": "string",
"doi": "string",
"key_findings": ["string"]
}
],
"coverage": {
"total_entries": "int",
"matched_entries": "int",
"gap_areas": ["string"]
}
},
"execution_handler": "sub-knowledge-updater",
"timeout": 10,
"retry": 1
}
Hooks & Lifecycle Management
Hook Types
The skill implements clean, reusable hooks for lifecycle events:
Pre-Execution Hooks
class PreExecutionHooks:
"""Hooks invoked before skill execution"""
@hook("language_detection")
def detect_language(context: ExecutionContext) -> LanguageConfig:
"""Detect user language and configure translation"""
# Auto-detect from query + context
# Return language configuration
@hook("requirement_validation")
def validate_requirements(requirements: Dict) -> ValidationResult:
"""Validate structured requirements before proceeding"""
# Check completeness, consistency
# Return validation result
@hook("tool_availability")
def check_tools_availability(tools: List[str]) -> AvailabilityResult:
"""Verify required tools are accessible"""
# Check API status, authentication
# Return availability + fallback options
Execution Hooks
class ExecutionHooks:
"""Hooks invoked during skill execution"""
@hook("before_tool_call")
def before_tool_call(tool_name: str, params: Dict) -> ModifiedParams:
"""Modify parameters before tool execution"""
# Add defaults, validate, transform
# Return modified parameters
@hook("after_tool_call")
def after_tool_call(tool_name: str, result: Any) -> ProcessedResult:
"""Process tool output before use"""
# Validate, extract, transform
# Return processed result
@hook("on_tool_failure")
def on_tool_failure(tool_name: str, error: Exception) -> FallbackResult:
"""Handle tool failure gracefully"""
# Log, attempt fallback, degrade gracefully
# Return fallback result or raise
Post-Execution Hooks
class PostExecutionHooks:
"""Hooks invoked after skill execution"""
@hook("quality_gate_validation")
def validate_quality_gates(output: Dict) -> GateResult:
"""Validate output against quality gates"""
# Check all gate criteria
# Return gate results + auto-fix suggestions
@hook("evidence_verification")
def verify_evidence_chain(evidence: List[Dict]) -> VerificationResult:
"""Verify evidence chain integrity"""
# Check citations, tiers, traceability
# Return verification result
@hook("output_formatting")
def format_output(raw_output: Dict, template: str) -> FormattedOutput:
"""Apply output template and formatting"""
# Structure according to template
# Return formatted output
Hook Execution Flow
User Query
↓
[Pre-Execution Hooks]
├─ language_detection
├─ requirement_validation
└─ tool_availability
↓
[Sub-Skill Execution]
├─ before_tool_call (per tool)
├─ tool execution
├─ after_tool_call (per tool)
└─ on_tool_failure (if needed)
↓
[Post-Execution Hooks]
├─ quality_gate_validation
├─ evidence_verification
└─ output_formatting
↓
Final Output
Modular Directory Structure
Directory Purpose & Contents
308-red-tide-algal-bloom-monitoring/
│
├── skills/ # Sub-skill definitions
│ ├── main.md # Primary harness orchestrator
│ ├── sub-gather-requirements.md
│ ├── sub-evidence-collector.md
│ ├── sub-core-analysis.md
│ ├── sub-knowledge-updater.md
│ └── sub-advisor.md
│
├── scripts/ # Automation & setup routines
│ ├── setup.py # Environment initialization
│ ├── data_ingestion.py # Bulk data import
│ ├── batch_analysis.py # Batch processing
│ └── export_results.py # Result export utilities
│
├── references/ # Domain knowledge & templates
│ ├── hab_species_reference.md
│ ├── toxin_profiles.md
│ ├── detection_methods.md
│ ├── prompt_templates.md
│ └── quality_standards.md
│
├── assets/ # Static resources
│ ├── diagrams/ # System & process diagrams
│ ├── schemas/ # JSON schemas for validation
│ └── templates/ # Output templates
│
├── config/ # Configuration management
│ ├── default_config.yaml # Base configuration
│ ├── local_config.yaml # Environment overrides (gitignored)
│ ├── config_loader.py # Type-safe config loader
│ └── schema.py # Configuration validation schemas
│
├── tools/ # Python utilities
│ ├── knowledge_updater.py # Knowledge crawl pipeline
│ ├── test_knowledge_updater.py
│ └── run_test_scenarios.py
│
├── tests/ # Test suites
│ ├── test-scenarios.md
│ └── TEST_RESULTS.md
│
├── logs/ # Structured logging output
│ └── skill.log # Main log file (auto-generated)
│
└── docs/ # Additional documentation
├── API.md # API documentation
└── DEPLOYMENT.md # Deployment guide
Graceful Degradation Protocol
Degradation Levels
The skill implements 5-level graceful degradation:
DEGRADATION_LEVELS = {
0: "Full Functionality",
1: "Reduced Data Availability - Use cached data + KB",
2: "Knowledge Base Only - No external sources",
3: "Simplified Analysis - Basic methods only",
4: "Explicit Limitation Flag - Minimal viable output"
}
Degradation Triggers
DEGRADATION_TRIGGERS = {
"satellite_api_failure": 1,
"sensor_network_unavailable": 1,
"authoritative_sources_timeout": 2,
"knowledge_base_corrupted": 3,
"all_external_sources_failed": 4,
"llm_rate_limit": 2,
"context_window_exceeded": 3
}
Degradation Banners
Level 1: ⚠️ LIMITED DATA AVAILABILITY - Using cached data and knowledge base. Some satellite/sensor data unavailable.
Level 2: ⚠️ REDUCED FUNCTIONALITY - External sources unavailable. Analysis based on knowledge base only.
Level 3: ⚠️ SIMPLIFIED ANALYSIS - Full analysis unavailable. Providing basic assessment with reduced detail.
Level 4: ⚠️ MINIMUM FUNCTIONALITY - Multiple system failures. Output requires verification with authoritative sources.
Quality Gates
Universal Gates (U1-U6)
| Gate | Criterion | Auto-Fix | Exit Condition |
|---|---|---|---|
| U1 | ≥3 sources cited, ≥1 academic/authoritative | Search missing sources | Retry max 2 |
| U2 | Safety/risk/limitation disclosure present | Add disclosure section | Retry max 2 |
| U3 | Evidence hierarchy stated per source | Label tiers explicitly | Retry max 1 |
| U4 | Language matches user preference | Translate if mismatched | Retry max 1 |
| U5 | Output uses project template | Restructure per template | Retry max 2 |
| U6 | Every claim traceable to source OR flagged as judgment | Add attribution flags | Retry max 2 |
Domain Gates (G1-G4)
| Gate | Criterion | Auto-Fix | Exit Condition |
|---|---|---|---|
| G1 | Bloom detection uses ≥2 methods (satellite + in-situ) | Add fallback method | Retry max 1 |
| G2 | Toxin risk cites regulatory limits explicitly | Add regulatory reference | Retry max 1 |
| G3 | Aquaculture impact includes species-specific vulnerability | Add species detail | Retry max 1 |
| G4 | Alert level matches defined categories | Map to standard category | Retry max 1 |
Performance Optimization
Context Management
class ContextManager:
"""Manage context window and token consumption"""
def optimize_context(self, messages: List[Dict]) -> List[Dict]:
"""Optimize context for token efficiency"""
# 1. Remove redundant content
# 2. Compress repetitive information
# 3. Prioritize recent/relevant messages
# 4. Apply caching for repeated access
def estimate_tokens(self, text: str) -> int:
"""Estimate token count for text"""
# Use model-specific tokenizer
def apply_compression(self, text: str, target_ratio: float) -> str:
"""Compress text to target ratio"""
# Preserve critical information
# Remove redundancy
# Maintain readability
Caching Strategy
class CacheManager:
"""Manage multi-level caching"""
def get(self, key: str, level: CacheLevel) -> Optional[Any]:
"""Get cached value by level"""
# L1: In-memory cache (session)
# L2: File cache (persistent)
# L3: Knowledge base fallback
def set(self, key: str, value: Any, ttl: int, level: CacheLevel):
"""Set cached value with TTL"""
def invalidate(self, pattern: str):
"""Invalidate cache entries matching pattern"""
Structured Logging
import logging
from typing import Dict, Any
class StructuredLogger:
"""Structured logging with context"""
def __init__(self, name: str, config: LoggingConfig):
self.logger = logging.getLogger(name)
self.config = config
def log(self, level: str, event: str, **kwargs):
"""Log structured event"""
log_entry = {
"timestamp": datetime.utcnow().isoformat(),
"level": level,
"event": event,
"skill": "red-tide-algal-bloom-monitoring",
"context": kwargs
}
self.logger.log(getattr(logging, level), json.dumps(log_entry))
def log_tool_call(self, tool: str, params: Dict, result: Any, duration_ms: int):
"""Log tool execution"""
self.log("INFO", "tool_call",
tool=tool,
params=params,
duration_ms=duration_ms)
def log_quality_gate(self, gate: str, passed: bool, details: str):
"""Log quality gate check"""
self.log("INFO", "quality_gate",
gate=gate,
passed=passed,
details=details)
Execution Example
Input
Analyze red tide risk for oyster farms in Apalachicola Bay, Florida for the next week
Internal Processing
# 1. Language Detection (English)
# 2. Requirements Gathering
requirements = {
"object_of_analysis": "red tide risk",
"scope": "oyster farms",
"region": "Apalachicola Bay, Florida",
"timeframe": "next 7 days",
"target_audience": "aquaculture",
"language": "en"
}
# 3. Evidence Collection
evidence = {
"satellite_chl_a": {"mean": 12.5, "max": 25.3, "anomaly": +3.2},
"sst": {"mean": 28.5, "anomaly": +1.5},
"in_situ_counts": {"Karenia_brevis": 8500, "total": 12000},
"authoritative_docs": ["NOAA HAB-OFS alert", "Florida FWC report"]
}
# 4. Core Analysis
analysis = {
"detection": {"bloom_status": "Active", "confidence": "High"},
"triggers": {"temperature": "Elevated", "stratification": "Present"},
"toxin_risk": {"risk_level": "Moderate", "toxin_type": ["Brevetoxins"]},
"aquaculture_impact": {"species_affected": ["oysters"], "vulnerability": "Moderate-High"}
}
# 5. Knowledge Base
knowledge = [
{"citation": "Blondeau-Patissier et al. 2014", "tier": 1},
{"citation": "Stumpf et al. 2003", "tier": 1},
{"citation": "IOC-UNESCO HAB guidelines", "tier": 1}
]
# 6. Synthesis
conclusion = {"category": "Monitor (elevated)", "confidence": "High"}
Output
# Red Tide Risk Assessment: Apalachicola Bay Oyster Farms
**Analysis Date**: 2025-01-16
**Risk Category**: Monitor (elevated)
**Confidence**: High
## Current Status
- **Bloom Status**: Active bloom detected
- **Cell Concentration**: 8,500 Karenia brevis cells/L
- **Satellite Chl-a**: 12.5 mg/m³ (anomaly: +3.2 mg/m³)
- **SST**: 28.5°C (anomaly: +1.5°C)
## Environmental Conditions
- **Temperature**: Elevated (+1.5°C above normal)
- **Stratification**: Present (favorable for bloom maintenance)
- **Nutrients**: Moderate conditions
- **Hydrodynamics**: Low circulation in bay
## Toxin Risk Assessment
- **Toxin Type**: Brevetoxins (PbTx)
- **Risk Level**: Moderate
- **Regulatory Status**: Below closure threshold (800 μg/kg)
- **Trend**: Increasing over past 48h
## Aquaculture Impact
- **Affected Species**: Oysters (Crassostrea virginica)
- **Vulnerability**: Moderate-High
- **Accumulation Rate**: Moderate (current cell count)
- **Depuration Time**: 7-14 days if harvest closure occurs
## Recommended Actions
1. **Monitor Daily**: Increase cell count monitoring to daily
2. **Pre-emptive Harvest**: Consider harvest before bloom peak (48-72h)
3. **Sample Testing**: Conduct toxin testing twice weekly
4. **Contingency Planning**: Prepare for potential harvest closure
## Scenarios
- **Most Likely (60%)**: Bloom continues for 5-7 days, moderate toxin accumulation, harvest closure recommended
- **Best Case (25%)**: Bloom dissipates within 72h, minimal impact
- **Worst Case (15%)**: Bloom intensifies, rapid toxin accumulation, immediate closure required
## Evidence Chain
1. Satellite chl-a anomaly → NASA MODIS/VIIRS (2025-01-16)
2. Karenia brevis cell counts → Florida FWC monitoring (2025-01-15)
3. Toxicity threshold → NOAA HAB-OFS standards (Tier 1)
4. Oyster vulnerability → Aquaculture research (Tier 2)
## Limitations & Uncertainties
**Limitations**:
- Satellite data may be affected by cloud cover
- In-situ measurements limited to 3 monitoring stations
- Toxin levels based on cell count correlations, not direct measurements
**Uncertainties**:
- Bloom trajectory dependent on wind patterns over next 72h
- Toxin accumulation rates vary with oyster physiological state
- Potential transport of bloom from adjacent waters
**Data Quality**: Good (multi-source, recent, consistent)
**Assumptions**:
- Current hydrological conditions persist for 48-72h
- No significant nutrient input events expected
- Karenia brevis remains dominant species
---
*Generated by red-tide-algal-bloom-monitoring v2.0.0*
*Processing time: 3.2s | Degradation level: 0 (Full functionality)*
Integration Points
Knowledge Pipeline
- Crawl Schedule: Weekly academic (Mondays 08:00) + daily news (07:00)
- Crawl Targets: ArXiv, Semantic Scholar, Harmful Algae, Remote Sensing journals
- Deduplication: SHA256 of DOI/URL
- Scoring: Composite (recency 0.4 + relevance 0.4 + citations 0.2)
External APIs
- Copernicus Marine Service: Satellite chl-a, SST, ocean color
- NOAA HAB-OFS: Operational forecasts, bulletins
- IOC-UNESCO HAB Portal: Global reports, species database
- Local Sensor Networks: In-situ fluorometers, water quality stations
Tool Dependencies
- WebSearch: Live domain news and reports
- WebFetch: Scrape authoritative sources
- Read/Write: Knowledge base and configuration access
- Bash: Python script execution (knowledge updater, tests)
Version History
- v2.0.0 (2025-01-16): Production-grade architecture with modular directories, hooks & tools, SKILL.md registry
- v1.0.0 (2024-12-01): Initial production-ready release with 6-phase harness
For implementation details, see PROJECT-detail.md and skills/*.md
For quality procedures, see tests/test-scenarios.md
For knowledge sources, see SECOND-KNOWLEDGE-BRAIN.md