Skill UI

A comprehensive framework for rigorously testing and benchmarking LLM agents. It moves beyond simple pass/fail checks to assess complex behaviors, reliability metrics, and capability consistency across multiple runs. Ideal for production monitoring and identifying subtle failure modes in advanced AI agents.

Fetch execution traces from LangSmith Studio to inspect LangChain/LangGraph agent runs, analyze errors, tool calls, memory usage, and performance, and export sessions for deeper diagnostics using the langsmith-fetch CLI.

This guide provides comprehensive strategies for managing AI agent resource consumption and credit usage. It details optimizing task steps, setting trigger filters, implementing webhook rate limiting, and setting up budget alerts. Use this guide to ensure efficient, cost-effective AI operations and prevent unexpected billing overruns.

Runs AgentShield to audit Claude Code configuration files and MCP components for secrets, injection risks, and overly permissive settings, offering CLI, JSON/Markdown/HTML reports, auto-fix, Opus deep analysis, and CI-ready GitHub Action hooks to harden new or existing projects.

Runs Semgrep static analysis with auto language detection, parallel Task subagents, Semgrep Pro checks, and configurable “run all” or “important only” modes to audit codebases for vulnerabilities and bug patterns.

Runs the Centinela QA agent to scan for dead code, tech debt, outdated or vulnerable dependencies, code smells, and TODO/FIXME audit results, generating a documented report for release or refactor readiness.

A comprehensive domain orchestrator designed to manage the entire project delivery lifecycle. It can route specific project management inquiries (e.g., Jira permissions, sprint velocity) to specialized sub-skills, or execute complex, goal-oriented workflows through a gated agentic loop. This process pulls live data from Jira/Atlassian, runs deterministic analytics, and verifies every step to ensure auditability and successful closure.

Autonomous experiment loop that optimizes any file by a measurable metric. The agent edits a target file, runs a fixed evaluation, keeps improvements (git commit), discards failures (git reset), and loops indefinitely. Ideal for optimizing code speed, reducing bundle size, improving test pass rates, enhancing prompts, or boosting content quality through iterative measured attempts.

Build a production-ready, AI-powered data collection agent for any public source (job boards, prices, news, etc.). This agent automatically scrapes data, enriches it using free LLMs (like Gemini Flash), and stores the results in databases such as Notion, Sheets, or Supabase. It runs on a free schedule using GitHub Actions and learns from user feedback, making it ideal for continuous data monitoring and tracking.

A foresight agent that runs a rigorous 12-step pipeline combining Claude’s qualitative intelligence with Python’s deterministic arithmetic to analyze high-stakes future scenarios, infer reasonable horizons, and surface quantitative signals for prediction requests.

Use this skill to deeply inspect and audit the entire process of local AI coding-agent sessions. It tracks critical operational metrics such as token consumption, cost spikes, tool failures, latency anomalies, and overall health scores. Ideal for debugging unreliable code runs, establishing CI/CD health gates, or comparing different attempts to ensure semantic accuracy.

This skill autonomously optimizes artifacts like code or models using Hypothesis Tree Refinement. It runs iterative experiments with dev/test evaluators to prevent overfitting, coordinating subagents in isolated worktrees. Ideal for improving training recipes, agents, or pipelines without manual supervision.