Ecological Cemetery Planning

Ecological Cemetery Planning Skill Registry

Overview

This document serves as the comprehensive registry for the ecological-cemetery-planning skill system. It defines how skills are registered, resolved, executed, and validated within the production-grade agent architecture.

Architecture Philosophy

The ecological-cemetery-planning skill implements a modular, event-driven architecture with:

• Hierarchical skill composition - Main orchestrator with specialized sub-skills
• Lifecycle hooks - Before/after/error handlers for each execution phase
• Tool registry - Dynamic tool loading with schema validation
• State management - Centralized state with change notification
• Quality gates - Multi-level validation with auto-fix and graceful degradation
• Knowledge pipeline - Self-improving crawl system with evidence hierarchy

Skill Registration

Core Skill Structure

ecological-cemetery-planning/
├── SKILL.md                    # This file - skill registry and documentation
├── CLAUDE.md                   # Skill identity and harness flow
├── PROJECT-detail.md          # Full technical specification
├── PROJECT-DEVELOPMENT-PHASE-TRACKING.md
├── README.md                   # Public-facing documentation
├── SECOND-KNOWLEDGE-BRAIN.md   # Living knowledge base
├── skills/
│   ├── main.md               # Main orchestrator skill (entry point)
│   ├── sub-gather-requirements.md
│   ├── sub-evidence-collector.md
│   ├── sub-core-analysis.md
│   ├── sub-knowledge-updater.md
│   └── sub-advisor.md
├── tools/
│   ├── knowledge_updater.py   # Knowledge crawl pipeline
│   ├── validate_project.py   # Project validation suite
│   ├── test_knowledge_updater.py
│   └── run_test_scenarios.py
├── config/
│   ├── config.py             # Configuration management system
│   ├── config.schema.json    # JSON schema for configuration
│   ├── default.json          # Default configuration values
│   └── hooks_system.py       # Hooks and tools registry
├── scripts/
│   ├── setup.py             # Environment setup and initialization
│   ├── seed_knowledge.py    # Initial knowledge base seeding
│   ├── run_validators.py    # Run all validation suites
│   └── deploy.py            # Production deployment script
├── references/
│   ├── domain-methods.md    # Authoritative domain methodology references
│   ├── evidence-tiers.md    # Evidence hierarchy and tier definitions
│   ├── quality-gates.md     # Complete quality gate specifications
│   ├── templates.md         # Output templates and formats
│   └── sources.md          # Curated domain sources with access methods
├── assets/
│   ├── diagrams/           # System architecture and flow diagrams
│   ├── schemas/            # JSON schemas for validation
│   └── examples/           # Example inputs and outputs
└── tests/
    ├── test-scenarios.md   # Concrete test scenarios
    └── TEST_RESULTS.md     # Test execution results

Skill Metadata Schema

Each skill file must include YAML frontmatter:

---
name: {skill-or-subskill-name}
description: {one-line summary of when to use this skill}
version: {semantic-version}
author: {author-info}
license: {license-type}
compatibility: {platforms-supported}
---

Skill Resolution Process

When a user invokes /ecological-cemetery-planning, the skill resolution follows:

1. Trigger Detection - Claude matches user query against skill description
2. Skill Loading - Load SKILL.md metadata (~100 words)
3. Main Skill Activation - Load skills/main.md full content
4. Sub-Skill Resolution - Dynamically load sub-skills as needed
5. Resource Loading - Load references/, assets/ files on-demand
6. Execution - Run harness through all phases with hooks
7. Quality Gate - Validate output against all gates
8. Output Delivery - Return formatted, evidenced result

Execution Flow

Phase 1: Intake (sub-gather-requirements)

Input: Raw user message + optional attachments

Process:

1. Parse user intent and constraints
2. Identify object of analysis (site, project, concept)
3. Clarify scope (full planning vs. specific question)
4. Determine timeframe and available inputs
5. Identify target audience and language preference

Output: Structured requirements object

{
  "object_of_analysis": "string",
  "scope": "full_project|specific_question|comparison",
  "timeframe": "string",
  "available_inputs": ["list"],
  "target_audience": "practitioner|researcher|decision_maker|learner",
  "language": "en|vi",
  "analysis_type": "planning|assessment|design|compliance"
}

Quality Gate: At least one object of analysis confirmed

Phase 2: Evidence Collection (sub-evidence-collector)

Input: Requirements object from Phase 1

Process:

1. Search real-time domain news and reports
2. Fetch authoritative documents and standards
3. Retrieve reference benchmarks from knowledge base
4. Collect site-specific data if provided
5. Identify relevant regulatory frameworks

Output: Evidence bundle

{
  "current_data": {
    "source": "string",
    "date": "ISO8601",
    "content": "object"
  },
  "authoritative_docs": [
    {
      "title": "string",
      "source": "string",
      "url": "uri",
      "date": "ISO8601",
      "tier": "Tier 1|2|3|4"
    }
  ],
  "recent_news": ["list"],
  "reference_benchmarks": ["list"]
}

Quality Gate: Current data + ≥1 authoritative doc retrieved, OR limitation flag

Phase 3: Core Analysis (sub-core-analysis)

Input: Evidence bundle + requirements

Process:

1. Site Assessment - Analyze soil conditions, hydrology, ecology
2. Burial Criteria - Apply green burial standards (Green Burial Council)
3. Landscape Design - Native plant cultural landscape plan
4. Capacity Planning - Calculate sustainable burial capacity
5. Maintenance Planning - Design sustainable maintenance protocol
6. Scenario Analysis - Compare development scenarios

Output: Analysis scorecard

{
  "site_assessment": {
    "soil_impact": "low|medium|high",
    "groundwater_risk": "none|buffer|risk",
    "ecological_value": "low|medium|high",
    "constraints": ["list"]
  },
  "green_burial_criteria": {
    "meets_standards": "boolean",
    "gaps": ["list"],
    "certification_path": "string"
  },
  "landscape_design": {
    "native_plant_communities": ["list"],
    "cultural_elements": ["list"],
    "ecological_restoration": "string",
    "maintenance_plan": "string"
  },
  "capacity_analysis": {
    "total_capacity": "number",
    "annual_capacity": "number",
    "sustainability_years": "number",
    "expansion_options": ["list"]
  },
  "scenarios": [
    {
      "name": "string",
      "description": "string",
      "pros": ["list"],
      "cons": ["list"],
      "impact_score": "number"
    }
  ]
}

Quality Gate: All analysis sections completed with specific findings

Phase 4: Knowledge Update (sub-knowledge-updater)

Input: Topic keywords from current analysis

Process:

1. Query SECOND-KNOWLEDGE-BRAIN.md for relevant entries
2. Surface authoritative citations with tier labels
3. Identify knowledge gaps for crawl pipeline
4. Optionally supplement with real-time search (≤2 queries)

Output: Knowledge citations

{
  "citations": [
    {
      "title": "string",
      "authors": ["list"],
      "year": "number",
      "doi": "string",
      "tier": "Tier 1|2|3|4",
      "relevance": "high|medium|low",
      "excerpt": "string"
    }
  ],
  "coverage_rating": "comprehensive|adequate|limited|gaps_identified",
  "gaps": ["list"],
  "crawl_recommendations": ["list"]
}

Quality Gate: ≥1 academic/authoritative source surfaced; coverage rating provided

Phase 5: Advisory Synthesis (sub-advisor)

Input: Core analysis scorecard + evidence bundle + knowledge citations

Process:

1. Synthesize findings - Integrate all analysis components
2. Evaluate scenarios - Compare against criteria
3. Assess risks - Identify environmental, regulatory, operational risks
4. Form recommendation - Select optimal approach with rationale
5. Plan remediation - Identify required actions and timeline
6. Structure disclosure - Explicit risk/limitation disclosure

Output: Final advisory

{
  "conclusion": "Ecologically Sound Plan|Conditional (buffer required)|Groundwater Risk|Inconclusive",
  "recommendation": "string",
  "key_risks": [
    {
      "risk": "string",
      "severity": "low|medium|high",
      "mitigation": "string"
    }
  ],
  "evidence_chain": {
    "claim": ["list"],
    "sources": ["list"],
    "confidence": "high|medium|low"
  },
  "remediation_steps": [
    {
      "action": "string",
      "priority": "immediate|short|medium|long",
      "responsible": "string",
      "timeline": "string"
    }
  ],
  "mandatory_disclosure": "string"
}

Quality Gate: Conclusion is exactly one of four categories; disclosure appears before conclusion

Phase 6: Quality Gate (main.md)

Input: Draft advisory output

Process:

1. Evidence Hierarchy Check - Verify all claims cited to Tier 1-3 sources
2. Disclosure Check - Confirm risk/limitation disclosure present
3. Format Check - Verify all mandatory sections present
4. Language Check - Confirm output matches user language preference
5. Traceability Check - Verify every claim traceable to source
6. Auto-Fix - Attempt to fix missing sections/citations
7. Retry - If auto-fix fails, retry up to 2 times
8. Degradation - If still failing, degrade with explicit limitation banner

Output: Validated final advisory

Quality Gates: All U1-U6 universal gates + G1-G4 domain gates must pass

Quality Gates Specification

Universal Gates (U1-U6)

| Gate | Criterion | Auto-Fix | Max Retry | |------|-----------|----------|-----------| | U1 | ≥3 sources cited, ≥1 academic/authoritative | Search for missing | 2 | | U2 | Safety/risk/limitation disclosure present | Add disclosure statement | 2 | | U3 | Evidence hierarchy stated per source | Label sources | 2 | | U4 | Language matches user preference | Translate output | 1 | | U5 | Output uses declared template | Restructure to template | 2 | | U6 | Every claim traceable to source OR flagged as judgment | Add flags | 2 |

Domain Gates (G1-G4)

| Gate | Criterion | Auto-Fix | Max Retry | |------|-----------|----------|-----------| | G1 | Soil/water impact assessed with specific findings | Re-run analysis | 2 | | G2 | Green-burial criteria applied with gaps identified | Re-run analysis | 2 | | G3 | Native-plant landscape design with species list | Re-run analysis | 2 | | G4 | Capacity/maintenance plan with sustainability metrics | Re-run analysis | 2 |

Graceful Degradation Levels

| Level | Trigger | Output | User Message | |-------|---------|--------|--------------| | 0 | All gates pass | Full validated advisory | None | | 1 | Minor gaps (≤2 gates) | Advisory with limitation notes | "Partial results: some gaps filled with defaults" | | 2 | Major gaps (3-4 gates) | Limited advisory + recommendations | "Limited results: knowledge base insufficient" | | 3 | Critical gaps (5+ gates) | Framework + manual guidance | "Framework provided: requires manual completion" | | 4 | System failure | Fallback to general guidance | "Service degraded: using general ecological guidance" |

Input/Output JSON Schemas

Input Schema

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "title": "Ecological Cemetery Planning Input",
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "User's natural language query or task description"
    },
    "site_data": {
      "type": "object",
      "properties": {
        "location": {"type": "string"},
        "area_hectares": {"type": "number"},
        "soil_type": {"type": "string"},
        "water_table_depth": {"type": "number"},
        "current_vegetation": {"type": "array", "items": {"type": "string"}},
        "existing_features": {"type": "array", "items": {"type": "string"}}
      }
    },
    "requirements": {
      "type": "object",
      "properties": {
        "burial_demand": {"type": "number"},
        "timeframe_years": {"type": "number"},
        "regulatory_context": {"type": "string"},
        "budget_constraints": {"type": "string"},
        "stakeholder_concerns": {"type": "array", "items": {"type": "string"}}
      }
    },
    "preferences": {
      "type": "object",
      "properties": {
        "language": {"type": "string", "enum": ["en", "vi"]},
        "output_format": {"type": "string", "enum": ["full_report", "summary", "recommendations_only"]},
        "detail_level": {"type": "string", "enum": ["high", "medium", "low"]}
      }
    }
  }
}

Output Schema

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "title": "Ecological Cemetery Planning Output",
  "type": "object",
  "required": ["metadata", "executive_summary", "analysis", "recommendations", "evidence_chain", "disclosure"],
  "properties": {
    "metadata": {
      "type": "object",
      "properties": {
        "execution_id": {"type": "string"},
        "timestamp": {"type": "string", "format": "date-time"},
        "version": {"type": "string"},
        "language": {"type": "string"},
        "degradation_level": {"type": "integer", "minimum": 0, "maximum": 4}
      }
    },
    "executive_summary": {
      "type": "object",
      "properties": {
        "conclusion": {"type": "string"},
        "key_findings": {"type": "array", "items": {"type": "string"}},
        "critical_recommendations": {"type": "array", "items": {"type": "string"}}
      }
    },
    "analysis": {
      "type": "object",
      "properties": {
        "site_assessment": {"type": "object"},
        "burial_criteria": {"type": "object"},
        "landscape_design": {"type": "object"},
        "capacity_analysis": {"type": "object"},
        "scenarios": {"type": "array", "items": {"type": "object"}}
      }
    },
    "recommendations": {
      "type": "object",
      "properties": {
        "primary": {"type": "string"},
        "remediation_steps": {"type": "array", "items": {"type": "object"}},
        "next_actions": {"type": "array", "items": {"type": "string"}}
      }
    },
    "evidence_chain": {
      "type": "object",
      "properties": {
        "sources": {"type": "array", "items": {"type": "object"}},
        "citations": {"type": "array", "items": {"type": "string"}},
        "confidence_assessment": {"type": "string"}
      }
    },
    "disclosure": {
      "type": "object",
      "properties": {
        "limitations": {"type": "array", "items": {"type": "string"}},
        "assumptions": {"type": "array", "items": {"type": "string"}},
        "risks": {"type": "array", "items": {"type": "object"}},
        "data_quality": {"type": "string"}
      }
    }
  }
}

Tool Registry

Available Tools

The following tools are available for use within the skill system:

WebSearchTool

• Category: DATA_FETCH
• Purpose: Search web for current information
• Input: {query: string, max_results: number}
• Output: {results: [{title, url, snippet, date}]}

WebFetchTool

• Category: DATA_FETCH
• Purpose: Fetch content from specific URL
• Input: {url: string, selector?: string}
• Output: {content: string, metadata: object}

KnowledgeQueryTool

• Category: KNOWLEDGE
• Purpose: Query knowledge base for relevant entries
• Input: {keywords: [string], tier?: string, limit?: number}
• Output: {entries: [{title, authors, year, doi, tier, excerpt}]}

ValidationTool

• Category: VALIDATION
• Purpose: Validate output against quality gates
• Input: {output: object, gates: [string]}
• Output: {passed: boolean, failures: [{gate, reason}], auto_fixed: [string]}

AnalysisTool

• Category: ANALYSIS
• **Pu

…

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: dungnotnull
• Source: dungnotnull/ecological-cemetery-planning-agent-skill
• License: MIT

Install and usage instructions live in the source repository linked above.