Skill UI

An expert skill for transforming CPU-bound Python code into highly performant GPU-accelerated code using NVIDIA technologies. It specializes in optimizing scientific workflows, including machine learning (ML), physics simulations, geospatial analysis, and large-scale array computations. Utilize tools like CuPy for array operations, Numba CUDA for custom kernels, and Warp for spatial computing and differentiable simulations, achieving dramatic speedups (10x to 1000x). Use this when dealing with NumPy, pandas, SciPy, or complex computational loops that benefit from parallel processing on NVI...

RDKit is a powerful, open-source cheminformatics library providing extensive Python APIs for molecular analysis and structural manipulation. It supports molecular I/O (SMILES, SDF, InChI), calculating physicochemical descriptors (MW, LogP), generating fingerprints, performing substructure searching, and analyzing stereochemistry. This toolkit is essential for advanced research in drug discovery, computational chemistry, and materials science.

A comprehensive debug bundle designed to collect diagnostic evidence from AppFolio property management API integrations. It captures essential data points including API connectivity status for properties, tenants, and work orders; authentication status using client credentials; recent integration error logs; and environment metadata. This allows support engineers to perform root cause analysis and diagnose issues like connectivity failures or authentication rejections without needing live environment access.

This skill provides a comprehensive observability stack for Flexport logistics integrations. It sets up Prometheus metrics for API health tracking, uses structured logging (pino) for debugging, implements OpenTelemetry tracing for latency analysis, and defines actionable Grafana dashboards and alert rules. This ensures reliable, performant, and fully monitored logistics operations.

A comprehensive guide detailing advanced offensive fuzzing methodologies. It covers the full lifecycle from target identification (parsers, network protocols, kernels) to tool selection (AFL++, libFuzzer, Honggfuzz, syzkaller), building harnesses, curating seed corpora, executing fuzz campaigns, and finally, performing crash triage and vulnerability analysis. Essential for security researchers and embedded systems developers.

Provides comprehensive observability for Salesforce integrations, covering API limit monitoring, SOQL performance tracking, error rate analysis, and structured logging. Use this skill when building robust, reliable connections to Salesforce, ensuring you track API consumption, detect performance bottlenecks, and set up critical alerting using tools like Prometheus and Grafana.

A hands-on methodology for testing IoT and embedded devices, covering hardware reconnaissance, firmware extraction and analysis, bootloader access, runtime exploitation, wireless protocol attacks, companion app review, and cloud API abuse.

A practical offensive methodology for Android and iOS app testing. It covers static analysis, Frida/Objection instrumentation, SSL pinning bypass, exported component abuse, insecure storage review, and mobile API reconnaissance for pentest and bug bounty work.

Covers WPA3/SAE attack techniques including transition-mode downgrade, Dragonblood side-channel analysis, SAE authentication flooding, H2E checks, and Wi-Fi 6E implications. Useful for assessing wireless security gaps and validating defensive hardening.

This tool empowers users to query, filter, and aggregate logs from Datadog directly from the command line using the Composio CLI. It bypasses the graphical UI, providing structured JSON output for complex analysis, cross-service correlation, and automated reporting. Ideal for deep dive investigations, tracking latency regressions, or building scheduled operational digests.

This meta-skill implements a combinator pattern for deep security analysis. It runs three independent gates—policy governance, secret scanning, and audit logging—over a candidate operation. It enforces a strict arbitration rule (Policy DENY > Scanner WARN > ALLOW) to provide a robust, multi-layered verdict, making it ideal for reviewing critical code changes or environment modifications for safety compliance.

This meta-skill is designed to automate the diagnosis of complex programming errors. When presented with a stack trace, traceback, or failing log, it orchestrates a multi-step investigation. It parses the trace, searches the repository history, checks relevant issues, and analyzes code diffs to provide a comprehensive brief for root cause analysis and debugging.