claudemods

Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).

38 skills

Brainstorm, plan, debug, review, and compound learnings with AI agents

37 skills

Production-grade engineering skills covering every phase of software development: spec, plan, build, verify, review, and ship.

25 skills · 4 agents

Persistent memory system for Claude Code - context compression across sessions

22 skills · 1 MCP server · hooks

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

Frontend design taste skills including brutalist, minimalist, soft, redesign, stitch, and more

13 skills

Editorial diagram design: 42 diagram types as self-contained HTML and SVG, with Mermaid, draw.io and Excalidraw import.

1 skill · 6 commands

Lossless context management — DAG-based summarization that preserves every message

4 skills · 1 MCP server · hooks

Forces the laziest solution that works. YAGNI, stdlib first, one line over fifty.

6 skills

Claude-Mem for Cowork (Claude app cloud sessions) - hooks stream tool use to cmem.ai and inject observations into new sessions and agents

3 skills · hooks

Colorful, themeable rendering for Claude Code replies: tables, headings, code, numbers, paths and links.

4 skills

btop-style live dashboard for Claude Code internals: /cctop opens it in a side pane, cctop-insights lets the session answer questions from its numbers

2 skills

Action-first, ADHD-friendly output shaping for Claude Code.

1 skill · hooks

Sentry error tracking and performance monitoring - search issues, analyze errors, explore traces, and get AI-powered root cause analysis

1 agent · 1 MCP server

Sentry MCP with experimental features enabled - includes all stable tools plus bleeding-edge capabilities

1 agent · 1 MCP server

Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.

1 MCP server

Rewrite AI-sounding text so it reads like the writer without changing what it says.

1 skill

29 production-ready scripts for iOS app testing, building, and automation

1 skill

A skill that plans, writes, validates, proves, reviews, migrates and publishes Claude Code mods. Plan first with the smallest $ surface, a footprint diffed against the plan, a test seen failing, an isolated load that leaves evidence files, a five-line threat model, and a handoff that names what was and was not proven. Brainstorm mode checks ideas against every mod already on GitHub.

1 skill

Agent Skill for general-purpose browser automation with Playwright. The agent writes and executes custom scripts for testing pages, validating UX, and automating browser tasks.

1 skill

Interactive setup wizard for theme, style, segment, and budget configuration

1 command

/flow opens a live, read-only tree of the session's subagents and teammates beside the transcript: status, current tool call, waits for approval, elapsed time, tokens; prints the same tree as text where no pane can be drawn

Watch the agent work: every step in plain English, with a live time-saved counter

Race Claude Code sessions on a task: live split-screen race track scoreboard

Read-only side chat: /aside opens a pane where you ask about the session so far, answered by a tool-less fork of the transcript; nothing is written back into the main thread.

Failing tests spawn a pixel boss; every run that fixes tests lands a hit, zero failures is a KO with loot

Stops a session before it runs past a budget: warns near a cost or plan-window limit, then refuses tool calls and asks before the next prompt.

A pane beside the transcript that holds the next things to tell the agent: write blocks of text while a turn runs, then press [+] to put one block into the prompt box.

Live session cost odometer above the prompt: a burning fuse, real-world comparisons, threshold alerts

cache-tax, the Mod. Stops a cold send once with its price, keeps the 1-hour prompt cache warm for a window you set (/keepwarm 6h), arms that for you after a paid cold write, and keeps score. Reach L2, one $.model.fork per idle stretch, its own store, a status slot; sees prompt length only.

Games above the Claude Code prompt (snake, Tetris, Doom, 2048, Minesweeper, Flappy, Pong, typing test, Space Invaders) that pause when Claude finishes a turn, and a pet that grows as Claude tests, commits and edits. Clones of the genre, not affiliated with or endorsed by Tetris Holding, Taito, Atari or id Software

Switches the Chrome profile Claude drives with no approval click.

A pixel Clawd in a side pane that reacts to what Claude is doing

A pixel pet that lives in a pane and reacts to what Claude does: eats on tool calls, gets sick on failures, panics on rm -rf, sleeps when idle, evolves as you ship

Live /context: what fills the window, per-turn growth, and distance to auto-compaction.

Click-to-copy buttons above the prompt for the last answer's blocks, plus a durable stash.

A band above the prompt: session and last-message cost, plus git branch, status, CI, pull and push

Verifies the live URL after a deploy command succeeds.

A docked pane with each edited file's hunk and keep or revert buttons that act without a model turn.

Blinks the iTerm2 tab orange when a turn lands, until you type again or three minutes pass.

Pins every subagent to the fable model by rewriting agent.spawn, with /fable-pin on, off and status. Reach L0: its own store and one command, nothing else.

A pixel-art companion above the prompt that reacts to your session, levels up, and can be redrawn by hand

Preview pasted images at their prompt markers on macOS. Ghostty, plus experimental iTerm2 and Herdr support with a startup override. Reach L2: local clipboard capture and temporary files, no model or network calls.

Gmail + Slack alerts inside Claude Code: toasts, status count, and an Alerts pane

A pane of Claude's dry inner thoughts about your session, typed out live

Dangerous Bash commands need launch codes: red alert pane, siren, a code to arm and a LAUNCH to fire

Paints Markdown onto the prompt box as you type: syntax-highlighted code blocks, inline code, bold, italic, lists, tables, quotes, links and headings. Paint only, your text is never changed.

Denies a merge unless the latest human message says merge; ship, push and deploy do not count.

Mermaid diagrams drawn as box art inline in the transcript.

Guided breathing exercises above the Claude Code prompt while Claude works; the spinner reads the breath

Routes heavy Bash commands to a second machine over ssh, syncing the commit first.

A money line above the prompt, read from SQLite databases over ssh.

Animated pixel-art scenes or your own image above the prompt, reacting to Claude. Zero tokens.

/q <text> while Claude is working waits in a stack above the prompt and goes out when the turn ends.

A rail of your session's prompts beside the transcript or above the prompt; hover to read one, click to jump to it.

A pane beside the transcript with the GitHub pull requests and issues related to the current session: drag a title or description to leave a comment on it, add as many as you like, then Submit sends them all as one prompt. Watch each pull request's checks, review decision and mergeability update on a timer.

5-hour and weekly plan limits with reset countdown and projected time until you hit them.

A receipt in the turn footer: edits, runs and curls, a warning when edits ran nothing, unfolded destructive commands, and a spinner that names an unverified claim.

claude review: the session's latest reply in a calm reading pane docked beside the conversation

Side pane showing the files changed in this conversation as diffs; mark lines, comment, and send the comments back to Claude

Movie-style credits for your Claude Code session

Stops a turn that has edited too many files until the goal is restated.

Keeps secrets, email addresses and IP addresses out of the session transcript, and puts the real value back on the way into a tool call

A side panel with live cards for every open Claude Code session. Requires CLAUDE_CODE_ENABLE_FUNCTION_HOOKS=1.

Spotify Wrapped for a Claude Code session: an animated stat reveal and a shareable PNG card

Live TV play-by-play of your Claude Code session, spoken aloud with crowd effects

Mission control for subagents and agent teams: who is running, what each is doing, who spawned whom, who is talking to whom

Per-turn cost, tokens and cache hit ratio, with the session total.

Keep unrelated topics from pulling Claude off your code: listed terms become stable placeholder names (or their lines are dropped) before the model reads them.

Limit windows, context fill and cost above the prompt, with a nudge to clear.

A usage board for Claude Code: a weekly pace gauge and a sliding ticker of your spend per model, in the terminal and on the desktop.

Pins the Vercel deploy queue of the linked project under the prompt: every deploy queued, building or just finished, with its phase and elapsed time

A pixel Ragtag athlete above the prompt: every turn is a rep, the session is an AMRAP.

3, 2, 1, GO on submit, a running clock per turn, and the split when it lands.