dotfiles

Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.

Pre-push Claude-driven code review loop with auto-fix. Runs parallel review agents (CLAUDE.md, bugs, git history, code comments, security, test coverage, and — on substantial diffs — structural simplification and observability coverage; plus, in cycle 1 on feature work, an intent-reconciliation agent that models what the PR should do from its purpose and reconciles that against the code to catch omission bugs) per cycle, scores each finding 0-100 with an independent Haiku rater, auto-fixes ≥80 findings, asks about 50-79 in a single batch per cycle, and stores per-repo learnings in .git/info/. Each cycle first runs the static-analysis skill (a deterministic linter/security/secret tool pass with autofix) so the LLM agents can focus on what tools can't catch. On clean exit, a final manual-testing evidence gate verifies the PR carries proof the changed functionality actually works — and produces that proof itself (local dev servers + playwright + surge-image-upload) when missing — reconciles the PR description to

Run every applicable static-analysis tool on a repo — detect languages/configs, run the curated CodeRabbit-weighted analyzer set (installed or ephemerally via npx/uvx), and write results to .static-analysis/. Use when asked to run static analysis, lint the whole repo, run all linters/analyzers, do a code-quality/security scan, or check a repo before pushing. Report-only by default; --fix opts into safe autofixers. Subagent-safe (never blocks).

Drive a pull request to a mergeable state: no conflicts with the base branch, green CI, and a clean review. Integrates the base branch and resolves conflicts, delegates PR feedback resolution (human + bot/CodeRabbit) to the resolve-feedback skill, runs the local review loop, fixes failing CI (via the fix-ci skill), and pushes. Waits once for CodeRabbit's initial automatic review to land and resolves its feedback, but does not re-request reviews or wait for CodeRabbit again after later pushes (CodeRabbit reviews each PR once, on open). Use when the user wants to drive a PR forward / get it to green / make it mergeable, resolve PR comments or review feedback, fix CI, or resolve merge conflicts — even if there is no feedback yet. Drives to mergeable but does NOT merge.

Retrieve, classify, and resolve PR review feedback — inline review threads, review summaries, and generic PR comments, from humans and bots. Fixes valid feedback (committing locally), posts justifications for invalid feedback, files follow-ups for out-of-scope items, and — critically — literally marks each item resolved/dismissed on GitHub (a reply is NOT a resolution). Supports an interactive mode (prompts on human feedback) and a non-interactive/batch mode (auto-applies bot/procedural feedback, defers human feedback to the caller). Use when the user wants to resolve PR feedback / review comments / review threads, address reviewer comments, or 'resolve feedback'. This is the feedback engine that the drive-pr and pr-triage skills delegate to; it does not push, run the review loop, fix CI, or integrate the base branch — that is the caller's job.

Audit and sync all pending changes across the dotfiles and dotprivate bare repos. Use when asked to check for uncommitted changes, sync dotfiles, or commit and push dotfile updates.

Generate a summary of GitHub activity for a configured client team. Shows PRs and issues each team member has been involved with. Default is last 7 days; specify a number of days to change the lookback period. Use when asked about team activity, what people have been working on, or to get a status update. Client configs live at ~/.claude/skills/client-report/clients/.

Analyze GitHub Actions usage for a GitHub org and identify ways to cut billable minutes. Estimates billable minutes per workflow (private repos only — public repos run free), ranks the biggest cost drivers, compares CI vs Copilot review, and recommends concrete efficiency fixes. Use when asked which CI checks use the most minutes, how to reduce Actions costs/usage, why CI won't start (budget/minutes block), or for a periodic Actions spend review. Accepts an org name.

Address failing CI checks for the PR associated with the currently-checked-out git branch. Use when asked to fix CI, address failing checks, fix the build, or investigate why CI is failing — including when checks fail to *start* (startup_failure, 0 jobs, 'workflow file issue'), which is often an account-level GitHub Actions budget/minutes block rather than a code problem (it diagnoses this before trying to fix code).

Audit my Claude Code scaffolding (skills, hooks, settings) for staleness and over-engineering against current Claude Code/Anthropic capabilities and my own logged friction, then propose deletions, simplifications, and fixes. Use when asked to run a bitter-lesson pass, audit my skills, find over-engineered or stale scaffolding, check what the harness now does natively, or do a periodic scaffolding upgrade.

Report CodeRabbit's current review status for the current PR (or a given PR). Use when the user asks where CodeRabbit is, whether CodeRabbit has finished reviewing, if CodeRabbit is rate limited/paused/still running, or for a quick CodeRabbit status check.

Go through open pull requests, check their status, and take actions to move them forward. This includes triaging PRs, fixing CI, resolving feedback, merging, or managing PR workflow. Use when asked to triage PRs, go through open PRs, or manage PR workflow. Runs autonomously by default on the user's own PRs — rebasing, resolving conflicts, fixing CI, and applying bot feedback without input — and only returns to the user once it has exhausted all autonomous work.

Groom project issues by reviewing the most stale (least recently touched) issues one at a time. Runs a priority-ordered series of checks on each issue, summarizes results, then walks the user through each failing check's fix one at a time for approval. Use when asked to groom issues, clean up the backlog, or review stale issues.

Scaffold a new Claude Code skill (personal or project-specific). Use when asked to create a skill, add a skill, or build a new slash command.

Go through open issues one-by-one, assess each for actionability, and spawn background agents to fix actionable ones in isolated worktrees with PRs. Use when asked to blitz through issues, batch-fix issues, or auto-resolve open issues.

Surface open Dependabot security alerts for the current GitHub repo, prioritized by severity and cross-referenced with any open Dependabot PRs. Use when asked to check Dependabot alerts, review security vulnerabilities, see what dependencies need patching, or triage what to address.

Turn colored terminal / ASCII output (charts, TUIs, CLI output, asciigraph, lipgloss/bubbletea views) into a PNG image — preserving ANSI colors — for embedding in PRs, issues, docs, or READMEs. Use when asked to screenshot a TUI/CLI, capture terminal output as an image, or "add a screenshot" of text-based/ASCII output. Pairs with the surge-image-upload skill to host the result and `gh` to embed it. Not for GUI/browser screenshots (use playwright).

Leverage local Ollama models (qwen2.5-coder) for the current task. Invoke when you want Claude Code to offload cheap, recoverable subtasks — bulk filtering/extraction/classification, or redacting secrets/PII before sending text externally — to a local model instead of doing them in-context.

Load Fieldnotes context and conventions. Use when starting work on any project or topic, when creating or updating project notes, or when the user asks about Fieldnotes structure or conventions.

Clean up stale git worktrees in the current repo — any tool (Claude, Zenflow, Vibe Kanban, PR triage, etc.) — by checking which branches are merged, then prompting the user to delete them. Use when asked to clean up worktrees, prune old worktrees, or tidy up worktrees from any tool.

Brainstorm and evaluate names for a product, company, or library against Daniel Reeves' nominology criteria, then check domain availability across many TLDs. Use when the user asks for naming help, says 'name it', 'help me name', 'naming brainstorm', or mentions 'nominology'.

Guides Stripe integration decisions — API selection (Checkout Sessions vs PaymentIntents), Connect platform setup (Accounts v2, controller properties), billing/subscriptions, Treasury financial accounts, integration surfaces (Checkout, Payment Element), migrating from deprecated Stripe APIs, and security best practices (API key management, restricted keys, webhooks, OAuth). Use when building, modifying, or reviewing any Stripe integration — including accepting payments, building marketplaces, integrating Stripe, processing payments, setting up subscriptions, creating connected accounts, or implementing secure key handling.

Prepare and compute tax forms using OpenTaxSolver (OTS) v23.06 for Tax Year 2025. Create and edit input files, run solvers, and generate filled PDF forms. Supports US 1040, Schedule C, Schedule SE, MI-1040, and other forms. Use when asked to prepare taxes, fill out tax forms, run tax calculations, or generate tax PDFs with OpenTaxSolver.

Interactively plan the day by reviewing open GitHub issues/PRs, recent time tracking, and having a conversation about priorities and what's on the user's mind. Produces a day plan appended to the Obsidian daily note. Use when asked to plan the day, figure out priorities, or organize today's work.

Structured, iterative workflow for executing complex, multi-day software development tasks and epics. Use this skill whenever a user wants to plan or execute a large development task, feature, epic, refactor, or project — especially when they mention ADHD, feeling overwhelmed, needing focus, wanting step-by-step guidance, or asking how to "break down" or "tackle" a complex coding problem. Also trigger when the user mentions PRs, pull requests, stacking PRs, GitButler, or wants to split work across multiple branches. This skill keeps cognitive load low, offers numbered menus at every decision point, integrates freeform feedback seamlessly, splits work into logical PRs (standalone or GitButler stacks) during decomposition, and writes a persistent progress file to the repo so work can resume across sessions. Trigger even if the user just says something like "I need to build X, where do I start?"

Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML/CSS layouts, or when styling/beautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.

Commit, push, branch, and manage version control with GitButler. Use for: commit my changes, check what changed, create a PR, push my branch, view diff, create branches, stage files, edit commit history, squash commits, amend commits, undo commits, pull requests, merge, stash work. Replaces git - use 'but' instead of git commit, git status, git push, git checkout, git add, git diff, git branch, git rebase, git stash, git merge. Covers all git, version control, and source control operations.

Find user-visible improvements (UVIs) deployed to production in the current git repo that are not hidden behind a feature flag. Use when asked to find recent improvements, list what's been deployed to prod, or review what's visible to users.

Review and improve an existing Claude Code skill. Use when asked to refine a skill, improve a skill, review a skill, or clean up a skill.

Interactively walk through a pull request diff hunk-by-hunk, requiring the user to prove understanding of each hunk before advancing. Use when asked to review a PR, walk through changes, or help assess a pull request.

Trim long idle/static periods in a screen recording (or any video) down to a max duration using ffmpeg scene detection. Use when asked to trim idle periods, remove dead air, or shorten a video recording.

Review the current conversation for frictions and problems, then update or create project skills to prevent them in the future. Use when finishing a task, after hitting repeated issues, or when asked to improve project skills.

Browser automation using playwright-cli. Use when asked to open a browser, navigate to a URL, click elements, fill forms, take screenshots, scrape web content, or automate browser interactions.

Safely split changes from the current branch into a separate pull request. Use when asked to "split this PR", "extract changes into a separate PR", "create a separate PR for X", or when a PR is too large and needs to be broken up.