Agent skill

tool-ecosystem-manager

Manages and improves the Bible study tool creation ecosystem. Use this skill when you need to change tool creation processes, reorganize tool directories, propagate learnings across tools, or analyze common issues. This is a meta-skill for improving the tool-experimenter and tool creation workflow itself.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/productivity/tool-ecosystem-manager-authenticwalk-context-grounded-bib

SKILL.md

Tool Ecosystem Manager

Overview

This skill manages the meta-level of Bible study tool creation - improving the processes, guidelines, and structures that create the tools themselves. While tool-experimenter improves individual tools, this skill improves the entire tool creation ecosystem.

When to Use This Skill

Use this skill when:

  • User wants to reorganize tool directories or change the tool structure
  • User wants to add new quality checks or validation rules to all tools
  • User wants to propagate learnings from one tool to others
  • User wants to analyze common issues across multiple tools
  • User wants to update the TEMPLATE, REVIEW-GUIDELINES, or other foundational documents
  • User wants to verify consistency across all tools
  • User wants to improve the tool-experimenter skill itself

Do NOT use this skill for:

  • Creating or improving a single Bible study tool (use tool-experimenter instead)
  • Running a tool to generate data (use bible-researcher instead)

Key Governing Documents

These documents control how tools are created and must be understood deeply:

Core Process Documents

  • @.claude/skills/tool-experimenter/SKILL.md - The skill that creates/improves individual tools
  • @bible-study-tools/TEMPLATE.md - Template for all tool README.md files
  • @REVIEW-GUIDELINES.md - Universal validation framework for all outputs
  • @SCHEMA.md - YAML structure standards for all data files
  • @STANDARDIZATION.md - Naming conventions for books, versions, languages, citations
  • @CLAUDE.md - Project overview and core principles

Tool-Specific Documents (per tool)

  • bible-study-tools/{tool-name}/README.md - Tool-specific instructions
  • bible-study-tools/{tool-name}/LEARNINGS.md - Accumulated knowledge from experiments
  • bible-study-tools/{tool-name}/experiments/ - Experimental data and revisions

Workflow Overview

Phase 1: Understand Request

  1. Determine if this is:
    • Process Change (Case 1): Reorganizing, adding validation, changing structure
    • Problem Analysis (Case 2): Analyzing issues, propagating learnings
  2. Create a git branch using gitflow syntax

Phase 2A: Handle Process Changes (Case 1)

This is for structural changes to how tools are organized or created.

Step 1: Analyze Current State

  1. Read all relevant governing documents
  2. Identify all tools that will be affected
  3. Map out dependencies and relationships
  4. Create a detailed change plan

Step 2: Update Governing Documents

Update documents in this order:

  1. REVIEW-GUIDELINES.md - If adding new validation criteria
  2. SCHEMA.md - If changing data structure
  3. TEMPLATE.md - If changing tool README template
  4. tool-experimenter/SKILL.md - If changing the tool creation workflow

Step 3: Update Existing Tools

For each affected tool:

  1. Update README.md to match new template/guidelines
  2. Update LEARNINGS.md to reflect new standards
  3. Move files if directory structure changed (when moving files git commit only the file changes so git will preserve the history)
  4. Ensure all examples still work

Step 4: Test Changes

  1. Select 2-3 representative tools
  2. Call bible-researcher subagent for each tool on test verses
  3. Verify outputs follow new guidelines
  4. Iterate if needed

Step 5: Document Changes

  1. Update version history in modified documents
  2. Create a summary of what changed and why
  3. Commit changes with clear message

Phase 2B: Analyze Problems and Propagate Learnings (Case 2)

This is for identifying common issues and spreading solutions.

Step 1: Gather Intelligence

  1. Read all LEARNINGS.md files from all tools
  2. Read sample YAML outputs from each tool's experiments and output dir
  3. Identify patterns:
    • Common problems mentioned in multiple LEARNINGS
    • Common solutions that worked well
    • Issues visible in YAML outputs (schema violations, missing citations, etc.)

Step 2: Categorize Issues

Organize findings by:

  • Universal issues - Affect all tools (→ update REVIEW-GUIDELINES or TEMPLATE)
  • Category-specific - Affect similar tools (→ document in relevant tool READMEs)
  • Tool-specific - Already documented in that tool's LEARNINGS
  • Resolved issues - Mark as solved in LEARNINGS so they're not worked on again

Step 3: Propagate Solutions

For each identified issue:

If Universal:

  1. Update REVIEW-GUIDELINES.md with new validation check
  2. Update TEMPLATE.md with guidance to prevent issue
  3. Add to tool-experimenter Phase 3 evaluation criteria

If Category-specific:

  1. Identify all tools in that category
  2. Add to each tool's README in "Common Challenges and Solutions" section
  3. Update LEARNINGS.md noting the solution was applied

If Already Solved in One Tool:

  1. Extract the solution from that tool's LEARNINGS
  2. Check if other tools have the same problem
  3. Apply solution and update their LEARNINGS

Step 4: Update LEARNINGS Files

For each tool's LEARNINGS.md:

  • Mark solved problems with [SOLVED - YYYY-MM-DD] prefix
  • Add cross-references to solutions used in other tools
  • Remove duplicate issues that are now documented centrally

Step 5: Analyze YAML Outputs

For output quality analysis:

  1. Read sample YAML files from experiments
  2. Check against SCHEMA.md standards
  3. Verify REVIEW-GUIDELINES compliance
  4. Identify patterns:
    • Missing citations
    • Schema violations
    • Fabricated data indicators
    • Incomplete word coverage
    • Generic statements

Document findings in:

  • Tool-specific LEARNINGS.md for unique issues
  • REVIEW-GUIDELINES.md for universal patterns

Common Use Cases

Use Case 1: Reorganize Tools by Type

Example: "Break bible-study-tools directory into subdirectories words/commentary/topics"

Workflow:

  1. Analyze current tool types:

    bash
    ls -la bible-study-tools/
    
  2. Categorize each tool:

    • words: original-language-words, test-word-meanings
    • commentary: (none yet)
    • topics: (none yet)
    • clusters: grouping-semantic-clusters
  3. Create new directory structure:

    bash
    mkdir -p bible-study-tools/words
    mkdir -p bible-study-tools/commentary
    mkdir -p bible-study-tools/topics
    mkdir -p bible-study-tools/clusters
    
  4. Move tools:

    bash
    git mv bible-study-tools/original-language-words bible-study-tools/words/
    git mv bible-study-tools/test-word-meanings bible-study-tools/words/
    git mv bible-study-tools/grouping-semantic-clusters bible-study-tools/clusters/
    

    Be sure to commit after you finish moving as git follows the history better when moves are isolated from any other edits.

  5. Update references in:

    • tool-experimenter/SKILL.md (file paths)
    • TEMPLATE.md (if it references paths)
    • Each tool's README.md (relative paths to TEMPLATE, GUIDELINES, etc.)
  6. Test with bible-researcher on 2 tools (one from each category)

  7. Document change in CLAUDE.md and commit

Use Case 2: Add Fair Use Verification to All Tools

Example: "Add process to verify fair use policy compliance"

Workflow:

  1. Read plan/policy/fair-use.md to understand requirements

  2. Update REVIEW-GUIDELINES.md:

    • Add fair use check to Level 1 CRITICAL validation
    • Define what constitutes compliance
    • Add examples of violations
  3. Update TEMPLATE.md:

    • Add fair use section to "Research Methodology"
    • Include citation format that ensures fair use
    • Add to verification checklist
  4. Update tool-experimenter/SKILL.md:

    • Add fair use verification to Phase 3 evaluation
  5. For each existing tool:

    • Update README.md with fair use guidance
    • Review existing experiments for compliance
    • Update LEARNINGS.md if violations found
  6. Test by running bible-researcher on 3 tools:

    • One word study tool
    • One semantic cluster tool
    • One future commentary tool
  7. Verify outputs include:

    • Source-language text as primary
    • Convergence grouping for translations
    • Comparative analysis for divergence
    • No programmatic reconstruction possible
  8. Document findings and commit

Use Case 3: Propagate "Helpful Websites" Learning

Example: "Tool A found BibleHub concordance very useful, apply to similar tools"

Workflow:

  1. Read tool A's LEARNINGS.md to understand what made it useful

  2. Identify similar tools that would benefit:

    • Check each tool's "Purpose" in README
    • Match tools doing similar research
  3. For each similar tool:

    • Read current README to see if they use this source
    • If not, add to "Required Sources" in Phase 1
    • Update LEARNINGS.md noting this was added from tool A
  4. Update TEMPLATE.md:

    • Add to suggested sources section
    • Include guidance on when to use it
  5. Test on each tool edited by calling bible-researcher

    • Verify the source adds value
    • Check quality of extracted data
  6. Document cross-learning in each tool's LEARNINGS

Use Case 4: Identify and Fix Common Schema Violations

Example: "Analyze YAML outputs for pattern errors"

Workflow:

  1. Collect sample YAML files from all tools:

    bash
    find bible -name "*.yaml" -type f
    
  2. For each file, check:

    • Verse reference format (BOOK.chapter.verse with zero-padding)
    • Inline citations present
    • No standalone translation fields
    • No sources: arrays in patterns
    • Word positions sequential
    • All required fields present
  3. Categorize violations by frequency and tool

  4. For common violations:

    • Update SCHEMA.md with clearer examples
    • Update REVIEW-GUIDELINES.md with detection patterns
    • Add to TEMPLATE.md as warnings
  5. For tool-specific violations:

    • Update that tool's README with examples
    • Add to LEARNINGS.md with solution
    • Mark as [ISSUE FOUND - YYYY-MM-DD] for tracking
  6. Re-run problematic tools:

    • Use updated README
    • Generate new YAML
    • Verify violation is fixed
    • Mark as [SOLVED - YYYY-MM-DD] in LEARNINGS

Decision Framework

When to Update Which Document

Update REVIEW-GUIDELINES.md when:

  • Issue affects output quality universally
  • New validation criterion needed
  • Persona perspective missing
  • Hallucination pattern discovered

Update SCHEMA.md when:

  • Data structure needs clarification
  • New field types emerging
  • Inconsistent naming across tools
  • Citation format issues

Update TEMPLATE.md when:

  • Tool creation process improved
  • Section organization better understood
  • Examples more helpful
  • Common tool patterns identified

Update tool-experimenter/SKILL.md when:

  • Experiment workflow improved
  • New phase needed
  • Evaluation criteria changed
  • Parallelization strategy refined

Update individual tool README when:

  • Tool-specific process learned
  • Stellar example discovered
  • Source found helpful for that tool
  • Schema refinement for that tool type

Quality Checks

After any ecosystem change, verify:

Document Consistency

  • All file paths in SKILL files are correct
  • Cross-references between documents are valid
  • Examples match current schema
  • Version numbers updated

Tool Integrity

  • All tools still in correct directories
  • READMEs reference correct relative paths
  • No broken links in documentation
  • Git history preserved

Validation Pipeline

  • REVIEW-GUIDELINES comprehensive
  • SCHEMA examples accurate
  • TEMPLATE reflects best practices
  • tool-experimenter workflow coherent

Test Coverage

  • At least 2 tools tested per change
  • Diverse tool types represented
  • Edge cases considered
  • Outputs validate against guidelines

Git Workflow

When making ecosystem changes:

  1. Branch Naming:

    • Process changes: feat/process-{description}
    • Guidelines updates: docs/guidelines-{description}
    • Directory restructure: refactor/structure-{description}
  2. Commit Strategy:

    • Group related changes into a single commit (one commit per logical task, not per file)
    • Write clear commit messages: docs: update REVIEW-GUIDELINES with fair use check
    • Reference issue/discussion if applicable
  3. Testing Before Commit:

    • Run bible-researcher on test cases
    • Verify outputs match expectations
    • Check for regressions
  4. Pull Request:

    • Summarize what changed and why
    • List all affected tools
    • Include test results
    • Link to relevant discussions

Monitoring and Maintenance

Output Format

After completing an ecosystem management task, provide:

Summary

  • What was requested
  • What was changed
  • Why it was changed
  • Impact scope (how many tools affected)

Changes Made

  • List of documents updated
  • List of tools modified
  • List of directories restructured
  • Commit messages used

Testing Results

  • Which tools were tested
  • Test verses used
  • Validation results
  • Any issues discovered

Next Steps

  • Follow-up tasks needed
  • Tools that need attention
  • Documentation to update
  • Monitoring to establish

Examples of Ecosystem Changes

Example 1: Added Copyright Compliance

Request: Add fair use verification to all tools

Changes:

  • Updated REVIEW-GUIDELINES.md Level 1 with copyright check
  • Updated TEMPLATE.md with fair use methodology section
  • Updated 3 existing tool READMEs
  • Added to tool-experimenter Phase 3 evaluation

Testing:

  • Tested on original-language-words: ✅ PASS
  • Tested on grouping-semantic-clusters: ✅ PASS
  • No violations found in existing outputs

Result: All future tools will verify copyright compliance


Example 2: Reorganized by Tool Type

Request: Organize tools into words/commentary/topics subdirectories

Changes:

  • Created new directory structure
  • Moved 4 tools to appropriate categories
  • Updated all relative paths in tool READMEs
  • Updated tool-experimenter references

Testing:

  • Verified bible-researcher can still find tools
  • Tested path resolution: ✅ PASS
  • Git history preserved: ✅ PASS

Result: Tool discovery now organized by type


Example 3: Propagated BibleHub Learning

Request: All word tools should use BibleHub concordance (learned from tool A)

Changes:

  • Added BibleHub to TEMPLATE suggested sources
  • Updated 2 word study tool READMEs
  • Cross-referenced in LEARNINGS.md files

Testing:

  • Re-ran original-language-words with BibleHub
  • Concordance data quality improved: ✅
  • No performance degradation: ✅

Result: Word tools now consistently use BibleHub


Skill Limitations

This skill does NOT:

  • Create new Bible study tools (use tool-experimenter)
  • Generate Bible data (use bible-researcher)
  • Review individual tool outputs (that's in tool-experimenter Phase 3)
  • Make theological judgments (stays meta-level)

This skill DOES:

  • Improve the tool creation process
  • Maintain documentation consistency
  • Propagate learnings across tools
  • Reorganize tool structure
  • Analyze patterns across tools
  • Update validation frameworks

Version: 1.0.0 Created: 2025-10-29 Last Updated: 2025-10-29 Maintained By: Context-Grounded Bible Project

Didn't find tool you were looking for?

Be as detailed as possible for better results