Agent Claude Code
Use when the user needs deep analysis of recent code changes for bugs, logic errors, or regressions — tracing execution paths across multiple files to find issues before they reach production. Distinct from code-reviewer (which reviews PRs); this agent hunts specific change sets for defects.
Agent Claude Code
Use proactively when the user proposes, reviews, or refactors cloud infrastructure — Terraform spanning AWS, GCP, or Azure; CI/CD pipelines; networking; IAM; observability. Evaluates portability, security, cost, and long-term operability.
Agent Claude Code
Use when the user asks for a thorough code review, wants a second opinion on a PR, or needs security and design issues surfaced before merging. Dispatch for review work that should be independent of the author's context.
Agent Claude Code
Use when the user needs a map of an unfamiliar codebase before making changes — entry points, module boundaries, key functions, and impact analysis for a proposed change. Dispatch before touching an area no one on the current thread understands.
Agent Claude Code
Use when the user is building or debugging a data pipeline, ETL process, data transformation, or data quality check — ingesting from APIs, files, or databases and loading to a warehouse or downstream store.
Agent Claude Code
Use when the user needs deep analysis of large document sets or multi-file cross-referencing — Teams/SharePoint hand-offs, client decks, financial models, and long PDFs. Dispatch for work that would blow out the main agent's context (20+ pages, 5+ files, multi-sheet Excel models).
Agent Claude Code
Use when the user has multiple independent work streams that can run simultaneously — spawns sub-agents per stream in git worktrees, coordinates execution respecting dependencies, and returns a single consolidated summary to the main thread.
Command Claude Code
Extract, summarize, or tabulate enterprise documents (.docx, .xlsx, .pdf, images, etc.) that the user has shared.
Command Claude Code
Work a production issue from observation to reproduction to root cause to fix.
Command Claude Code
Add or update docstrings and module-level documentation for a file or directory. Pass the target path as an argument.
Command Claude Code
Generate a test suite for a specific file, module, or function. Pass the target path as an argument.
Command Claude Code
Execute an approved feature plan end-to-end through PR.
Command Claude Code
Scaffold a new skill, command, agent, or Cursor rule with correct frontmatter and location.
Command Claude Code
Analyze an existing codebase and generate a project-level CLAUDE.md that tells Claude Code what the project is, how it's structured, and how to work effectively within it.
Command Claude Code
Decompose a feature or epic into parallel and sequential work streams, assign file ownership, and optionally dispatch the parallel-worker agent to execute them.
Command Claude Code
Produce an implementation plan for a feature before writing any code.
Command Claude Code
Structured review of a PR or branch diff — correctness, security, design, readability.
Command Claude Code
Run a security audit on the current branch — surfaces credential leaks, injection risks, overly broad permissions, and dependency vulnerabilities.
Settings file Claude Code
A small, highly efficient computer algebra system in written in C99.
Skill Claude CodeCodex
Use when the user references or attaches Microsoft Office files (.doc, .docx, .xls, .xlsx, .ppt, .pptx), PDFs, images (.png, .jpg, .jpeg), or CSV/JSON/YAML from Teams or SharePoint and asks to extract, summarize, tabulate, or cross-reference their contents. Ships a runnable extract.py — invoke that directly rather…
Skill Claude CodeCodex
Use when the user asks to integrate a third-party REST or GraphQL API — fetching data, authenticating, handling rate limits, retrying on failure, and wiring responses into the application.
Skill Claude CodeCodex
Use when the user reports a bug, shows a stack trace, says "something's broken", or asks for help debugging unexpected behavior. Covers reproduction, isolation, root cause, fix, and regression test — the full loop. Applies to any language or stack.
Skill Claude CodeCodex
Use when the user asks to add, alter, or drop database tables or columns, create a migration script, or run schema changes safely against a live database.
Skill Claude CodeCodex
Use when the user describes a rough feature idea and wants to clarify requirements, resolve open questions, and produce a clean brief before writing a PRD. Activates on phrases like "I want to build", "I have an idea for", "what if we added", "can we make a feature that".