dotfiles

Create or improve agent skills. Load when creating SKILL.md files, writing skill descriptions, or structuring skill content for OpenCode or Claude.

Build stateful AI agents using the Cloudflare Agents SDK. Load when creating agents with persistent state, scheduling, RPC, MCP servers, email handling, or streaming chat. Covers Agent class, AIChatAgent, state management, and Code Mode for reduced token usage.

Reviews code changes for bugs, security regressions, type-safety issues, and maintainability. Use when asked to review uncommitted changes, the last commit, a PR/MR, a diff, or specific changed files.

Performs security-focused differential review of code changes (PRs, commits, diffs). Load for security audit, regression review, auth/crypto/access-control changes, validation removal, blast-radius analysis, adversarial review, or comprehensive markdown reports. Use standard code-reviewer for ordinary quick reviews.

Create and review Cloudflare Durable Objects. Use when building stateful coordination (chat rooms, multiplayer games, booking systems), implementing RPC methods, SQLite storage, alarms, WebSockets, or reviewing DO code for best practices. Covers Workers integration, wrangler config, and testing with Vitest.

Manages GitLab merge requests, CI/CD pipelines, and issues via glab CLI. Load before running glab commands, creating MRs, debugging pipeline failures, checking CI status, or any GitLab operation. Triggers on "merge request", "pipeline", "CI failure", "glab", or when git remote contains "gitlab".

Runs durable, long-running objectives through Codex-style goal continuation. Load when the user invokes /goal, creates or resumes a persistent goal, updates an active objective, asks for goal status, or needs a completion/budget/blocked audit. Uses goal runtime tools when available and prevents premature handoff by continuing checkpoint work until a real stop rule applies.

Use when adding, refactoring, or reviewing logging code and callsites across a codebase. Ensures structured, queryable, context-rich logging.

Guides use of the minipro CLI with XGecu/TL866 programmers, especially T48 over USB. Load before explaining or running minipro commands for chip selection, connection checks, EEPROM/EPROM reads, binary writes, verification, readback comparison, and checksum confirmation. Includes AT28C64B, 27C64, Motronic 8 KB ROM, Intel HEX, raw binary, erase, blank-check, and pin/contact checks.

Guides Porsche 911 Carrera 3.2 Bosch Motronic DME troubleshooting, diagnostics, ROM decoding, injector/fueling checks, ignition checks, harness debugging, and conservative tuning. Load for 1984-1989 3.2 Carrera DME faults, no-starts, drivability issues, EPROM/map analysis, AFR/timing questions, DME relay or sensor diagnosis, injector pulse, spark, AFM, CHT, ICV, speed/reference sensor, and 28-pin chip work.

Manages PlanetScale databases via MCP server or pscale CLI. Load before running pscale commands, using PlanetScale MCP tools, debugging connection errors, or analyzing slow queries.

Reviews, improves, and iterates on system prompts and LLM instructions. Load when writing, rewriting, reviewing, or optimizing a system prompt, agent prompt, tool description, skill description, or any LLM instructions. Triggers on "prompt engineering", "improve this prompt", "make this more LLM-friendly", or any block of instructions to refine. Applies model-aware analysis to improve clarity, instruction hierarchy, eval coverage, and model compliance.

Audits a codebase for unnecessary complexity -- premature abstractions, deep nesting, inconsistent patterns, and indirection that hurts readability. Use when asked to simplify code, audit complexity, reduce cognitive load, or review maintainability. Triggers: simplify, complexity audit, too complex, hard to read, maintainability review.

Summarizes session work for commit messages or branch reviews. Queries the OpenCode server API to find and read sessions, extract major changes, bug fixes, challenges requiring multiple iterations, and code review fixes. Load before committing or when preparing to document completed work.

Analyzes web performance using Chrome DevTools MCP. Measures Core Web Vitals and related lab metrics (FCP, LCP, INP/TBT, CLS, Speed Index), identifies render-blocking resources, network dependency chains, layout shifts, caching issues, and accessibility gaps. Use when asked to audit, profile, debug, or optimize page load performance, Lighthouse scores, or site speed.

Cloudflare Workers CLI for deploying, developing, and managing Workers, KV, R2, D1, Vectorize, Hyperdrive, Workers AI, Containers, Queues, Workflows, Pipelines, Secrets Store, and Pages. Load before running wrangler commands, editing wrangler.jsonc, changing bindings, deploying, deleting resources, or handling Worker secrets.