dotfiles

Capture session learnings and save to skills, guidelines, or reference docs under ~/.claude/.

Orchestrate parallel claude -p sessions — bootstrap, launch, monitor, and converge. Works with any skill that produces manifest.json, item directories, and a runner script.

Create a request (pull request or merge request) or update an existing one following project conventions.

Resolve merge or rebase conflicts between branches.

Assess open PRs with unaddressed review comments and generate a parallel addressing script — produces manifest.json and let-it-rip.sh for address-request-comments execution.

Assess open work items and generate a parallel execution script — produces manifest.json and let-it-rip.sh for implement/clarify execution.

Analyze a file for refactoring opportunities and apply selected improvements.

Fetch and address request comments from a pull request (GitHub) or merge request (GitLab).

Extract independent changes from a feature branch into a new PR targeting main.

Split a commit with mixed changes into separate commits

Single-pass curation of learnings, guidelines, and skills — consolidate, reorganize, or prune. For exhaustive multi-sweep curation, use learnings:consolidate instead.

Distribute global learnings and guidelines from ~/.claude/ into the current project's .claude/ directory.

Clean up stale research worktrees whose branches have been merged or deleted.

Run end-of-session retrospective: compound learnings, then discuss skill performance, communication, and improvements.

Activate a named domain persona to set focus, priorities, and gotchas for the current session.

Load explore-repo scan artifacts into context and produce a concise repository brief for Q&A.

Deep-scan a repository to understand its structure, functionality, and documentation gaps.

Code review a pull request or merge request. Fetches diff, analyzes through active persona lens, posts review with inline comments. Detects previous reviews and handles re-review automatically. Use when the operator asks to review a PR, do a code review, review a request, or review changes.

Clean up local branches that have been merged into main.

Multi-persona team code review. Orchestrates parallel reviewer subagents, merges findings with signal-strength tags, surfaces disagreements via deliberation. Use when the operator asks for a team review, multi-persona review, or team code review.

Assess open PRs and generate a parallel review script — produces manifest.json and let-it-rip.sh for team-review-request execution.

Extract learnings from request history (GitHub PRs or GitLab MRs) in batches — review patterns, architectural decisions, conventions, and engineering insights from discussions and metadata.

Fetch and analyze a request's (PR or MR) diff and metadata, then enter interactive Q&A mode.

Analyze a large request (PR or MR) and propose how to split it into smaller, reviewable units.

Initialize an autonomous consolidation loop with worktree, output scaffolding, and pre-flight checks.

Run exhaustive multi-sweep curation of all learnings, skills, and guidelines — auto-applies high-confidence recommendations and batches others for approval.

Pull and merge Claude skills, learnings, and guidelines from a configured sync source into the current repo.

Run a security audit on one or more projects using parallel agents.

Analyze a plan for parallelization opportunities and produce a structured parallel plan with a dependency DAG. Use when the operator says "parallelize this plan" or wants to split work into concurrent streams.

Rebase a chain of stacked/dependent branches when main is updated.

Execute a structured parallel plan by launching concurrent subagents with DAG-based scheduling. Use when the operator references a parallel plan file to execute.

Evaluate consolidation run quality, handle review items, and prepare to merge or relaunch.

Load a research project's files into context and produce a concise brief for Q&A.

Compare duplicate research directories to determine which is superseded.

Initialize an iterative research project with spec and progress tracking.

Resume a completed or blocked iterative research loop by collecting answers and relaunching.