@obra

systematic-debugging provides a four-phase debugging methodology prioritizing root cause investigation over symptomatic fixes. emphasizes reproducibility, evidence gathering in multi-layer systems, and architectural questioning when multiple fixes fail.

obra-using-git-worktrees detects existing isolation, prioritizes native platform tools, and falls back to git worktree creation with project-local directory safety checks, dependency installation, and baseline test verification.

obra-systematic-debugging enforces root cause investigation before any fix attempt, using four mandatory phases (investigation, pattern analysis, hypothesis testing, implementation) with explicit failure modes and architectural bailout criteria.

test-driven-development enforces writing failing tests before implementation code, following red-green-refactor cycles. covers feature development, bug fixes, and refactoring with explicit rules against code-first workflows and detailed rationalization counters.

receiving-code-review establishes a verification-first workflow for evaluating code review feedback, emphasizing technical correctness over performative agreement and providing decision logic for when to implement, clarify, or pushback.

verification-before-completion enforces evidence-based completion claims by requiring fresh command execution and output inspection before any success assertion, with explicit red flags and common failure modes.

obra-receiving-code-review covers technical evaluation of feedback before implementation, emphasizing verification over performative agreement, with explicit patterns for unclear items, external reviewer skepticism, and pushback reasoning.

obra-subagent-driven-development coordinates multi-task execution by dispatching fresh subagents per task, reviewing each against spec and code quality, and running a final whole-branch review. it optimizes for session continuity and fast iteration without human-in-loop delays between tasks.

writing-plans decomposes multi-step specs into focused, testable tasks with exact file paths, code samples, and commands. enforces tdd workflow and requires plan review before execution.

dispatching-parallel-agents covers delegating independent tasks to isolated agents without shared context or sequential dependencies, maximizing investigation speed across unrelated problem domains.

obra-writing-plans guides creation of detailed implementation roadmaps from specs, with task decomposition, file mapping, and bite-sized step breakdowns. targets engineers new to a codebase by enforcing complete code samples and exact file paths.

getting-started-with-skills documents mandatory workflows for skill discovery and usage within the implexa environment, emphasizing read-before-use discipline, todowrite integration for checklists, and explicit announcement patterns to maintain process transparency.

requesting-code-review dispatches a focused code reviewer subagent with git context (base/head SHAs) to catch issues before merge, emphasizing early and frequent review across task-driven development workflows.

obra-executing-plans coordinates implementation of written plans across multiple sessions with explicit review gates, task verification, and mandatory handoff to finishing-a-development-branch skill upon completion.

obra-requesting-code-review dispatches a focused code reviewer subagent with clean git context separation, covering mandatory review points (task completion, features, merges) and optional scenarios with explicit feedback triage rules.

writing-skills teaches test-driven documentation for agent techniques by establishing baseline failures, writing minimal skill docs to fix violations, then refactoring to close loopholes. requires prior tdd literacy.

subagent-driven-development provides a pattern for executing multi-task implementation plans by spawning isolated agents per task with two-stage (spec then code quality) review gates, avoiding context bleed and maintaining coordinator focus.

obra-writing-skills teaches test-driven development applied to skill documentation, with emphasis on writing failing test cases before skill content and iterating through red-green-refactor cycles.

You MUST invoke this skill before saying "I don't know," guessing, or treating any topic as new, no matter how trivial the question seems. It supplements other…

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting…

Use when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via…

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent,…

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup

Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always

Use when implementing any feature or bugfix, before writing implementation code

Use when you need direct browser control - teaches Chrome DevTools Protocol for controlling existing browser sessions, multi-tab management, form automation,…

using-tmux-for-interactive-commands — an installable skill for AI agents, published by obra/superpowers-lab.

Use when auditing a codebase for semantic duplication - functions that do the same thing but have different names or implementations. Especially useful for…

Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Makes…