Code can look complete while missing real behavior. This skill helps Claude, Codex, and Claude Code test implementations against actual execution evidence.

Distributed agent deployment can be hard to coordinate and monitor. This skill provides a structured Flow Nexus process for swarm setup, workflow execution, metrics, and scaling.

ML training can fail through NaN loss, overfitting, slow convergence, or unstable gradients. This skill guides diagnosis, targeted fixes, validation, and reporting.

Debugging can stall when symptoms, causes, fixes, and tests are handled separately. This skill guides a five-phase workflow for diagnosis, repair, validation, and regression prevention.

Teams need consistent slash commands that route micro-skills without confusing syntax. This skill guides command design, validation, handler generation, testing, documentation, and deployment steps.

Creating reusable AI skills requires consistent files and metadata. This skill guides Claude Code and Codex through a structured template workflow.

Manual presentation production can be slow and inconsistent. This skill guides Claude, Codex, or Claude Code through structured PPTX generation and validation.

Sandboxed coding agents need clear file, network, and resource boundaries to reduce host exposure. This skill provides a structured workflow for defining policies, creating isolation scripts, testing controls, and documenting the security posture.

Agent workflows often lose context between stages. This skill shows how to pass outputs through memory or streams with validation steps.

Production API builds need clear roles, tests, and deployment gates. This skill organizes agents through planning, TDD implementation, validation, documentation, and blue-green release.

Teams often need a web UI to trigger and observe CLI workflows. This skill provides a bridge pattern using REST endpoints, WebSockets, and local CLI integration.

GitHub Actions workflows can become slow, fragile, and hard to secure. This skill guides Claude, Codex, and Claude Code through structured CI/CD design, tuning, and review.

Security reviews often miss issues spread across code, dependencies, runtime behavior, and secrets. This skill guides Claude, Codex, and Claude Code through a structured audit workflow with reporting gates.

Style drift makes code reviews slower and less consistent. This skill guides Claude, Codex, or Claude Code through scans, reports, safe fixes, and validation.

Ambiguous requests can lead to incorrect plans, wasted work, and missed constraints. This skill analyzes intent, asks focused questions, and produces a clear execution brief.

Skill libraries become hard to govern as coverage, overlap, and usage drift. This skill guides Claude, Codex, and Claude Code through structured portfolio analysis.

Performance work is hard when swarm metrics, profiles, and benchmarks are scattered. This skill gives Claude, Codex, and Claude Code a structured analysis workflow.

Teams lose time when they pick the wrong Claude Code interface for a task. This skill recommends Web, CLI, or a hybrid handoff workflow.

AI-assisted code changes can be hard to trust before merge. This skill helps teams score outputs, run verification checks, and connect quality gates to CI, rollback, and monitoring workflows.

Placeholder code hides product risk and can reach production with fake behavior. This skill guides Claude, Codex, or Claude Code through a structured audit that finds, prioritizes, and replaces theater with real implementations.