rohitg00

Stress-test a plan by walking its decision tree one question at a time. Use when the user wants to pressure-test a design before implementation.

Complete AI coding workflow system with orchestration patterns, 18 hook events, 5 agents, cross-agent support, reference guides, and searchable learnings. It works with Claude Code, Cursor, and 32+ agents.

Surface past learnings relevant to the current task before starting work. Searches correction history, recalls past mistakes, and applies prior patterns.

Enforce directory lockdown when active.

Track parallel work sessions and prevent confusion across multiple Claude Code instances. Each major step ends with a status line, and every question re-states the project, branch, and task.

Assign a Thoroughness Rating (1-10) to each decision point, always choosing the higher-rated option as AI reduces the cost of doing things properly to near-zero. It also includes scope checks to differentiate between contained and unbounded work.

Start, structure, and grow a persistent research wiki, indexed in pro-workflow's SQLite knowledge base. Each wiki consists of markdown pages with provenance and a shadow FTS5 index for easy recall.

Run quality gates, review staged changes for issues, and create a well-crafted conventional commit. Use when saying commit, git commit, save my changes, or ready to commit after making changes.

Queries pro-workflow wikis using SQLite FTS5 BM25 retrieval to return top-K passages with citations. Use this tool to answer questions covered by user wikis, or when drafting new wiki pages to avoid duplication.

An end-of-session ritual that audits changes, runs quality checks, captures learnings, and produces a session summary. Use this when saying "wrap up", "done for the day", "finish coding", or ending a coding session.

Audit connected MCP servers for token overhead, redundancy, and security. Use when sessions feel slow or before adding new MCPs.

Smart context compaction with state preservation, saving critical files, task progress, and working state before compaction and restoring them afterwards. Use before manual compact or when auto-compact triggers.

LLM-powered quality verification using prompt hooks. It validates commit messages, code patterns, and conventions using AI before allowing operations, setting up intelligent guardrails.

Produces a one-screen map of an unfamiliar codebase area, detailing entry points, modules, data flow, and callers. Designed for fifteen-second comprehension, it provides quick orientation.

Displays session analytics, learning patterns, correction trends, heatmaps, and productivity metrics, computed from project memory and session history. Use for stats, progress, coding history, or dashboard inquiries.

Removes AI-generated code slop, unnecessary comments, and over-engineering from the current branch diff, cleaning up boilerplate, simplifying abstractions, and stripping defensive code. Use this when cleaning up code, simplifying, removing boilerplate, or before committing.

Create and manage git worktrees for parallel coding sessions with zero dead time. Use when blocked on tests, builds, wanting to work on multiple branches, context switching, or exploring multiple approaches simultaneously.

Analyze permission denial patterns and generate optimized alwaysAllow and alwaysDeny rules. Use when permission prompts are slowing you down or after sessions with many denials.

This skill compiles structured literature surveys on AI/ML topics. An agent curates research bundles from public resources, and a chosen LLM then generates the survey as a markdown wiki page, including a full bibliography.

Generates a structured handoff document detailing current progress, open tasks, key decisions, and context for resuming work. Use when ending a session with phrases like continue later, save progress, session summary, or pick up where I left off.

Reduce token waste by 40-60% through anti-sycophancy rules, tool-call budgets, one-pass coding, task profiles, and read-before-write enforcement. Inspired by drona23/claude-token-efficient.

Automatically grows a pro-workflow wiki by running a budget-capped BFS research loop over pluggable source fetchers (web, arXiv, GitHub). It fetches sources, drafts wiki pages, dedupes claims, and enqueues follow-up seeds, halting on budget, depth, or convergence.

Renders a self-contained, S3-uploadable HTML viewer for a pro-workflow wiki, consolidating all content and full-text search into a single file without external dependencies. It's ideal for visual browsing, sharing current states, auditing research, or handing off knowledge bases.

Coordinate multiple Claude Code sessions as a team, with a lead and teammates using shared task lists, mailbox messaging, and file-lock claiming. It includes patterns for team sizing, task decomposition, and when to use teams versus sub-agents versus worktrees.

Auto-configures quality gates, hooks, and settings for a new project, detecting its type and setting up appropriate tooling. Use when onboarding a new codebase.

Decompose large-scale changes into independent units and spawn parallel agents in isolated worktrees. Use for migrations, refactors, codemods, and any change touching 10+ files with the same pattern.

Captures user-reported defects as durable GitHub issues, written in the project's domain language. It explores the codebase for context without leaking file paths or line numbers, ideal for conversational bug reports or explicit requests to file an issue.

This skill enables provider-agnostic multi-LLM deliberation, featuring independent responses, cross-model anonymized ranking, and chairman synthesis. It's used for multiple AI perspectives or consensus-building, with provider config from env and optional transcript persistence to a wiki.

Captures corrections or lessons as persistent learning rules, categorizing them by mistake and correction. These rules are stored, categorized, and retrieved for future sessions, activated after mistakes or when the user explicitly requests to remember something.

Track session costs, set budget alerts, and optimize token spend. Use to check costs mid-session or set spending limits.

Master the four operations of context engineering: Write, Select, Compress, Isolate. Manage token budgets, compaction strategies, and context partitioning to keep AI sessions sharp and efficient.

Optimize token usage and context management. Use when sessions feel slow, context is degraded, or you're running out of budget.

Configure file watching hooks to auto-react to config changes, env file updates, and dependency modifications. Use to set up reactive workflows.

Wire Commands, Agents, and Skills together for complex features. Use when building features that need research, planning, and implementation phases.