Skill UI

Exa MCP provides advanced neural search capabilities across the web, specialized code repositories (like GitHub), and corporate/personnel databases. Use it for deep background research, finding the latest news, debugging complex code patterns, or conducting thorough market and competitor analysis.

This skill helps developers and QA engineers transform vague, intermittent, or poorly documented bug reports into a minimal, reproducible, and evidence-backed brief. It guides the user to isolate the exact failure conditions, separate expected outcomes from actual errors, and reduce the issue to the smallest safe failing fixture, ensuring diagnosis occurs before premature repair.

This skill provides an in-depth, entity-level scorecard for clinical or biomedical Named Entity Recognition (NER) models. It calculates precision, recall, and F1 scores (both strict and partial match modes) using a user-supplied gold corpus. Key features include per-label error breakdown, a confusion matrix, and detailed False Negative/False Positive analysis, essential for debugging model performance in highly sensitive medical domains.

This skill facilitates cross-version symbol migration and binary diffing. When existing symbols or reverse engineering results from an old version need to be quickly mapped to a new binary version, this tool is used. It leverages LLMs to perform structured structural comparisons between corresponding functions in different versions, enabling the programmatic transfer of function names and offsets even when original PDB files are lost. Ideal for kernel debugging, application updates, and binary analysis.

This diagnostic skill audits the entire local setup of Claude Code against several known, complex cache and cost bugs (CC#40524, B4, B5). It checks critical components including version compatibility, sentinel placeholders, session resuming logic, and attribution headers. Use this tool to ensure your development environment is optimized, stable, and minimizes unexpected operational costs and functional regressions.

A specialized downstream skill for deep security analysis of JSON Web Tokens (JWT), JWS, and JWE. It guides the identification of vulnerabilities where token components (headers, claims, keying) lead to unauthorized identity assumption or privilege escalation. Use this to trace the full path from token supply to final application acceptance.

A specialized skill for analyzing and reproducing complex, timing-dependent bugs in concurrent systems. Use this to investigate race windows, ordering bugs, lock gaps, stale reads, or state inconsistencies where the final outcome depends critically on request timing or asynchronous execution order. Essential for deep security auditing and debugging of distributed transactions.

This specialized skill guides deep-dive analysis for binary-heavy challenges, including CTF, malware analysis, and advanced forensics. It covers reverse engineering binaries, unpacking samples, inspecting memory dumps, debugging crashes, and building exploit chains by correlating artifacts like files, memory, and network packets.

A specialized, downstream skill designed for forensic analysis within a confined sandbox environment, particularly for CTF challenges. It guides the user to inspect the full stack behavior, including API interactions, server routing, browser state (cookies, local storage), and asynchronous worker processes. Use this skill to trace complex, end-to-end request flows, debug authentication/session leaks, or diagnose discrepancies between frontend UI and backend enforcement.

This local developer tool inspects dependency manifests (requirements.txt, pyproject.toml, package.json) for surface-level issues like standard-library shadowing, unpinned dependencies, and conflicts. It provides plain-language explanations and suggested fixes, helping developers debug dependency rot before deeper installation attempts.

This tool systematically audits and scores defined AI agents to ensure they are correctly scoped, highly specific, and safe for programmatic orchestration. It checks for description ambiguity, over-permissive tool grants, and anti-patterns that break headless workflows. Use it when onboarding new agents, integrating agent fleets, or debugging inconsistent multi-agent selection.

This tool audits and validates all Claude Code hooks defined across your settings files. It checks for validity, performance safety, correct event scoping, and proper exit-code handling (blocking vs. non-blocking). Essential for initial setup, debugging hooks that fail to fire, or performing routine system hygiene passes to ensure your agent's tool flow is robust and reliable.