Shubhamsaboo/awesome-llm-apps

• **Archetype**: mega-skill-pack
• **Stars**: 105,284
• **Last pushed**: 2026-04-13
• **License**: Apache-2.0
• **Discovered**: 2026-04-13
• **Source**: backlog evaluation
• **Skills found**: 20+ domain skills

Summary

Comprehensive collection of LLM applications and agent skills with educational content. Contains 20+ high-quality domain skills (debugger, deep-research, academic-researcher, fullstack-developer, etc.) plus a sophisticated self-improving agent framework that automatically optimizes skills using multi-agent collaboration. Includes MCP agents, RAG tutorials, voice AI agents, and advanced AI agent frameworks.

Assessment

Very high transferable value. Individual skills contain detailed procedural knowledge with systematic workflows (debugging process, research methodologies, development patterns). The self-improving agent system implements an automated skill optimization methodology using ADK agents (Executor, Analyst, Mutator) that could be adapted for babysitter's process library. Skills follow proper frontmatter format with clear usage guidelines, systematic processes, and concrete examples. Much more sophisticated than typical "expert persona" skills.

Extraction Priority

• High
• Rationale: 105k+ stars with comprehensive skill collection containing extractable procedural knowledge. Self-improving agent methodology is particularly valuable for automated process optimization. Multiple domain processes and meta-methodology applicable to babysitter.

Skills Inventory

Processes

- Source: awesome_agent_skills/debugger/SKILL.md (debugging process section) - Placement: specializations/shared/systematic-debugging - Inputs/Outputs: Bug report → Root cause + fix + prevention - Complexity: moderate - Notes: Includes binary search, rubber duck debugging, logging strategies, common patterns

• **Systematic Debugging Process**: 6-phase debugging workflow (understand, gather, hypothesize, test, identify, fix/verify)

- Source: awesome_agent_skills/academic-researcher/SKILL.md - Placement: specializations/research/academic-research - Inputs/Outputs: Research question → Comprehensive literature review + analysis - Complexity: moderate - Notes: Citation management, source evaluation, synthesis techniques

• **Academic Research Methodology**: Structured approach to academic research and literature review

- Source: awesome_agent_skills/deep-research/SKILL.md - Placement: specializations/shared/deep-research - Inputs/Outputs: Research topic → Verified findings + comprehensive report - Complexity: moderate - Notes: Source validation, fact-checking, cross-referencing methodology

• **Deep Research Pipeline**: Multi-stage research process with source verification and synthesis

- Source: awesome_agent_skills/self-improving-agent-skills/README.md (optimization loop) - Placement: methodologies/automated-optimization - Inputs/Outputs: Agent skill → Optimized skill + improvement changelog - Complexity: complex - Notes: ADK-based, uses Executor/Analyst/Mutator agents, Pydantic schemas, iterative improvement

• **Self-Improving Agent Framework**: Automated skill optimization using multi-agent collaboration

- Source: awesome_agent_skills/project-planner/SKILL.md - Placement: specializations/shared/project-planning - Inputs/Outputs: Project requirements → Detailed project plan + timeline - Complexity: simple - Notes: Requirements gathering, timeline creation, risk assessment

• **Project Planning Workflow**: Structured approach to project planning and execution

Plugin Ideas

- What install.md would do: Set up ADK integration, configure Executor/Analyst/Mutator agents, create optimization pipeline for existing processes - Processes it would copy: automated-optimization methodology - Configs/hooks it would create: ADK configuration, evaluation criteria templates, optimization pipeline configs - Source evidence: Complete self-improving agent system with multi-agent collaboration framework

• **Agent Skill Optimizer**: Plugin implementing the self-improving agent framework for babysitter processes

- What install.md would do: Install debugging tools, code review checklists, development workflow processes, testing frameworks - Processes it would copy: systematic-debugging, code-review-process, fullstack-development - Configs/hooks it would create: Debug logging configs, code review templates, git hooks for quality gates - Source evidence: Comprehensive development skills with systematic workflows and concrete tooling

• **Development Toolkit**: Plugin combining debugging, code review, and development workflow processes

- What install.md would do: Set up research methodology processes, citation management tools, source verification workflows - Processes it would copy: academic-research, deep-research, technical-writing - Configs/hooks it would create: Research templates, citation formats, fact-checking workflows - Source evidence: Detailed research methodologies with structured approaches and validation techniques

• **Research Suite**: Plugin providing academic research and deep research capabilities

Harness Integration Ideas

N/A - This is not a harness framework repository.

Implicit Procedural Knowledge

- Source: Self-improving agent system architecture (Executor/Analyst/Mutator collaboration) - Placement: specializations/shared/multi-agent-optimization - Why codify: Reusable pattern for iterative improvement of any artifact type (processes, skills, documentation) - Sketch: Define evaluation criteria → Execute/score → Analyze failures → Apply targeted fixes → Repeat

• **Multi-Agent Optimization Pattern**: Systematic approach to improving artifacts using specialized agent roles

- Source: Debugger skill strategies section (binary search, logging, bisection methods) - Placement: specializations/shared/debugging-strategy - Why codify: Systematic approach to debugging tool selection improves efficiency and success rate - Sketch: Classify problem type → Select investigation strategy → Apply technique → Verify results → Document findings

• **Debugging Strategy Selection**: Process for choosing appropriate debugging techniques based on problem characteristics

- Source: Deep research methodology with fact-checking and verification workflows - Placement: specializations/research/source-validation - Why codify: Ensures research quality and prevents misinformation propagation - Sketch: Identify sources → Cross-reference claims → Verify credentials → Rate reliability → Document provenance

• **Research Source Validation**: Process for verifying and cross-referencing information sources

Library Mapping

Plugin Marketplace Mapping