Agent Skills by Lars Bell
Showing 30 vetted skills indexed across 1 GitHub repositories.
lebsralai-taking-actions
Automate AI workflows by enabling agents to call APIs and use tools.
lebsralai-serving-apis
Wrap DSPy programs with FastAPI to expose query and health check endpoints.
lebsralai-decomposing-tasks
Decompose failing complex AI tasks into reliable subtasks using structured templates.
lebsralai-coordinating-agents
Orchestrate multiple AI agents with a LangGraph supervisor and handoff patterns.
lebsralai-building-chatbots
Build memory-enabled chatbots with DSPy and LangGraph for multi-turn conversations.
lebsralai-summarizing
Summarize long content into concise, audience-aware formats with faithfulness checks.
lebsralai-switching-models
Switch AI providers or models while preserving model-agnostic workflows.
lebsralai-cutting-costs
Optimize DSPy pipeline model routing, caching, and prompts to reduce AI API costs.
lebsralai-fine-tuning
Fine-tune models on labeled datasets using BootstrapFinetune workflows.
lebsralai-parsing-data
Extract structured data from unstructured text using DSPy.
lebsralai-stopping-hallucinations
Ground AI outputs in retrieved sources with enforced citations and verification.
lebsralai-building-pipelines
Decompose complex AI tasks into modular multi-stage DSPy pipelines.
lebsralai-moderating-content
Classify user-generated content, score severity, and route moderation decisions.
lebsralai-scoring
Score work against defined rubrics with per-criterion and overall scores.
lebsralai-sorting
Classify and route text content into predefined categories.
lebsralai-testing-safety
Identify AI safety vulnerabilities through automated adversarial testing with DSPy.
lebsralai-following-rules
Enforce hard and soft constraints on AI outputs using DSPy Assert and Suggest.
lebsralai-monitoring
Monitor AI model performance, drift, and cost in production.
lebsralai-improving-accuracy
Define custom metrics, run evaluations, and apply optimization strategies like BootstrapFewShot.
lebsralai-reasoning
Identify planning tasks and apply DSPy reasoning strategies with structured plans.
lebsralai-searching-docs
Builds a DSPy search-and-answer pipeline for grounded document retrieval and question answering.
lebsralai-generating-data
Generate synthetic training data with a DSPy-based generate-filter-evaluate workflow.
lebsralai-tracing-requests
Trace and inspect AI request steps with OpenTelemetry instrumentation and DSPy inspection.
lebsralai-kickoff
Scaffold a complete DSPy AI feature project with standard files and templates.