agenticluke

Patterns for automated Claude Code loops, from simple step-by-step jobs to RFC-based multi-agent work.

An engineering model for teams that use AI agents to write a large amount of code.

Build and run clear tests for Claude Code tasks. Use this skill to set pass or fail rules, test task success, check for regressions, measure pass@k and pass^k, or compare prompts and model versions.

Run long-lived agent services with clear logs, strict access, safe updates, and fast recovery.

Checks if an agent follows a skill, rule, or agent file. It makes three prompt levels, runs the agent, maps tool calls to required steps, checks step order, and writes a full report.

Use this skill to stop harmful actions in live systems, limit file edits, and make self-run agents safer.

Scan Claude Code setup files for secrets, unsafe access, bad hooks, risky MCP servers, and prompt attacks. Use after setup changes, before a commit, when joining a project, or for a regular safety check.

Safety rules for AI agents with wallet or trade access. Covers prompt attacks, spend limits, checks before send, stop rules, MEV safety, and key care.

Regression test rules for AI-assisted work. Test API routes without a database. Add a test for each fixed bug. Check that sandbox and live paths return the same shape.

Run notifications as one ECC-native flow across GitHub, Linear, desktop alerts, hooks, and connected message tools. Use when the real problem is alert routing, duplicate alerts, missed handoffs, escalation, or inbox overload.

Review DeFi automated market maker (AMM) smart contracts for flash loan attacks, bad slippage checks, sandwich attacks, price changes, reentry, unsafe token calls, access flaws, and math errors. Use for swaps, liquidity adds, liquidity removes, minting, burning, fees, and pool price logic.

Build clean Android and Kotlin Multiplatform apps with clear modules, dependency rules, use cases, repositories, data sources, Room, SQLDelight, Ktor, Koin, or Hilt.

Creates Angular code and gives architecture help. Use for Angular projects, components, services, signals, linkedSignal, resource, forms, dependency injection, routing, SSR, ARIA, animation, styles, tests, CLI tools, and Angular MCP setup.

Let the user choose answer depth before you reply. Use when the user asks to control answer length, detail, token use, or depth. Triggers include "token budget," "token count," "token usage," "token limit," "response length," "answer depth," "short version," "brief answer," "detailed answer," "exhaustive answer," "TL;DR," a depth percent, or close forms in any language. Do not use when the user already set a depth for this chat, the answer is clearly one line, or "token" means an auth, session, API, or payment token.

Build private, offline AI features with Apple's FoundationModels framework. Use it for text, typed output, tools, and live updates.

Evidence-based audit of ECC jobs, hooks, connectors, MCP servers, and scripts. Use it to find what is live, broken, stale, repeated, or missing before making changes.

Use connected billing tools, such as Stripe, to help with subscriptions, refunds, failed payments, cancellations, billing portals, and plan reviews. Use this skill when a customer needs billing help or when a billing change may affect revenue.

ECC's proof-first workflow for revenue, prices, refunds, team billing, and billing rules. Use it for sales snapshots, price checks, duplicate charge reviews, or billing answers that must match the code.

Find real, report-ready security flaws in a code repo. Use for bounty work, responsible reports, and checks for remote attack paths. Focus on flaws an outside attacker can reach and use. Skip weak, local-only, test-only, and out-of-scope findings.

Use browser tools after a deploy to test page views, user actions, speed, and access needs.

A GAN-style work loop that keeps building and review separate to improve app quality.

Watch a live URL for bugs after a deploy, merge, fix, or package update.

Review Cisco IOS and IOS-XE show commands, config modes, wildcard masks, ACL placement, interface health, and safe change checks.

Clear coding rules for TypeScript, JavaScript, React, Node.js, APIs, tests, and code reviews. Use when you write, review, or improve code.

Create CodeTour `.tour` files for a clear, step-by-step walk through real files and lines. Use for onboarding, system tours, pull request reviews, root cause reviews, and reusable guides that explain how code works.

Study a new codebase and make a clear guide with its structure, key entry points, code rules, and a starter CLAUDE.md. Use when joining a project, explaining a repo, or setting up Claude Code for the first time.

Patterns for Compose Multiplatform and Jetpack Compose apps. Covers state, events, navigation, themes, speed, and platform UI.

Find the right local context for a sub-agent through short search, review, and refine loops.

Run a steady agent loop with clear quality checks, tests, stop rules, and safe recovery steps.

Helps with customs papers, HS and tariff codes, duty planning, free trade rules, denied-party checks, and trade law across the US, EU, UK, and Asia-Pacific. Use for imports, exports, customs audits, and penalty notices.

Call a four-voice council for unclear choices, tradeoffs, and go or no-go calls. Use when several sound paths exist and the user needs clear, fair disagreement before choosing.

Record key software design choices as short Architecture Decision Records. Capture the reason, options, tradeoffs, risks, and status so future developers know why the system was built this way.

Build or grow a design system with clear tokens, reusable parts, themes, responsive layouts, and access checks. Use for Figma, Storybook, UI libraries, and keeping design files in sync with code.

Run Django jobs in the background with Celery. Covers task queues, workers, retries, schedules, and Redis or RabbitMQ.

Process, convert, scan, read, edit, sign, and fill documents with the Nutrient DWS API. Supports PDF, DOCX, XLSX, PPTX, HTML, and image files.

Use clear C# and .NET patterns for safe, fast, and easy-to-change apps. Covers dependency injection, async code, EF Core, options, results, APIs, and tests.

Test C# and .NET code with xUnit, FluentAssertions, mocks, integration tests, and clear test structure.

Adversarial review with a fix loop. Two independent reviewers must both pass the work before it can ship.

Audit ECC Tools for high token cost, extra pull requests, quota bypass, premium model leaks, duplicate jobs, and costly retries. Use only for the sibling ECC-Tools repo.

Helps users find, choose, install, reset, and use current Everything Claude Code parts. Check the live repo files before you answer.

Install, update, or set up ECC in Claude Code, Codex, or Kimi. Use only the plugin, scope, and hook features that each tool supports.

Prevent Ethereum hash bugs in JavaScript and TypeScript. Node SHA3-256 is not Ethereum Keccak-256. Using it can break selectors, signatures, storage slots, and address creation with no warning.

Use an eval-first workflow, small tasks, risk checks, and cost-aware model choice for AI-led software work.

Evidence-first research for current facts, comparisons, profiles, and advice based on public sources and local context from the user.

Create images, video, and audio with the fal.ai MCP server. Use for text-to-image, image edits, text-to-video, image-to-video, speech, music, sound effects, and video-to-audio tasks.

Safe Dart and Flutter patterns for null safety, fixed state, async work, widgets, state tools, routes, HTTP, code generation, app layers, and tests.

Manage carrier groups, check carrier risk, run freight bids, compare rates, track service, build routing guides, assign freight, and plan carrier reviews. Use this skill for truckload, LTL, intermodal, and broker work.

Manage freight delays, damage, loss, shortages, refused loads, claims, fees, and carrier disputes. Use this skill to sort risk, save proof, meet claim dates, plan next steps, and write clear updates.

Test F# code with xUnit, FsUnit, Unquote, FsCheck, stubs, and .NET test tools.

Use the gget CLI or Python package to search genome data, find gene IDs, fetch gene details or sequences, run quick BLAST or BLAT searches, check enrichment, and save a clear record of each query.

Manage GitHub issues, pull requests, CI runs, releases, and security alerts with the gh CLI. Use when asked to check GitHub, triage issues, review pull requests, fix CI, prepare a release, or review alerts.

Manage Google Drive files, automate Sheets, work with Gmail, and organize Docs and Slides.

Protect health and personal data in health apps. Use for access rules, APIs, databases, logs, audits, storage, exports, and security reviews.

Turn a local Hermes operator flow into a clean ECC skill and release package. Use this when a Hermes flow should be shared without private files, secrets, names, account data, or local paths.

Build and review US health software that must protect PHI and follow HIPAA.

Plan and review safe home lab network changes, security checks, speed tests, and rollout steps.

Plan and set up a safe home lab network, including devices, IP ranges, Wi-Fi, DNS, and network groups.

Use this skill when the user asks to create, write, add, or set up Hookify rules, or needs help with Hookify rule patterns and syntax.

Use this skill to read Jira tickets, find issues with JQL, study needs, add comments, update fields, link work, or move issues to a new status. Use Jira tools through MCP when they are ready. Use Jira REST API v3 only when MCP is not ready.

Store, sort, sync, find, and remove duplicate knowledge across local files, Claude Code memory, MCP memory, data stores, GitHub, Linear, and Git repos. Use when the user asks to save to a knowledge base, import notes, sync knowledge, clean up copies, or find what is known about a topic.

Use Kotlin coroutines and Flow in Android and KMP. Covers safe scopes, parallel work, StateFlow, SharedFlow, errors, cancel rules, and tests.

Write and improve Kotlin tests with Kotest, MockK, coroutine test tools, property tests, Kover, and the red-green-refactor TDD cycle. Use for new Kotlin code, missing tests, test bugs, slow or flaky tests, and coverage setup.

Find and review Laravel packages with the LaraPlugins.io MCP server. Use it to search for packages, check package health, and check Laravel or PHP support.

Find, rank, and contact strong leads. Use agent-based signal scores, shared contact ranks, warm intro paths, source-based voice matching, and channel-specific drafts for email, LinkedIn, and X. Use when the user wants to find, screen, or contact high-value people.

Check interface errors, drops, CRC errors, link flaps, speed, duplex, and counter changes on routers, switches, and Linux hosts. Use for packet loss, slow links, unstable links, rising error alerts, or checks before hardware changes.

Build iOS 26 Liquid Glass views in SwiftUI, UIKit, and WidgetKit. Use blur, tint, light, touch effects, shared glass groups, and smooth shape changes.

A clear workflow for reviews of academic, medical, technical, and science sources. It covers search plans, source checks, evidence tables, synthesis, and citation checks.

Use current library, framework, and API docs through Context7 MCP. Use this skill for setup, config, API details, code examples, or questions about tools such as React, Next.js, and Prisma.

Evidence-first steps for ECC live messages. Use when the user wants to read texts or direct messages, find a recent one-time code, check a thread before a reply, or show which message source was checked.

Build LLM pipelines that control cost with model choice, budget checks, safe retries, and prompt caching. Use for batch jobs, mixed task sizes, or any app with a firm API budget.

Build a full lobster persona for an OpenClaw AI agent. Use user choices or a random draw to make a role, SOUL.md text, in-character rules, a name, and an avatar prompt. If an approved image skill is available, it may also make the avatar. Use this skill to make a new OpenClaw lobster persona or finish a rough one. Do not use it for small edits to an existing SOUL.md, roles for other agent systems, or agents that should have no persona. Triggers include: lobster soul, lobster character, OpenClaw soul, lobster persona, lobster NPC, lobster story, lobster SOUL, random lobster, draw a card, and gacha.

Evidence-based email work for ECC. Use it to sort a mailbox, write drafts, send mail, check the Sent folder, or follow up safely.

Build clear, reusable Manim videos for technical ideas, charts, workflows, system maps, metrics, and product demos. Use when motion and structure matter more than real-life footage. For full video edits, pass the Manim render to the wider ECC video tools.

Modern Perl 5.36+ patterns for safe, clear, and easy-to-test apps and modules.

Build a new API connector or provider that matches the target repo's current design. Use this skill when adding one or more links to outside APIs without making a second integration system.

Review and improve a user's X and LinkedIn network. Build safe lists of people to keep, review, unfollow, follow, or contact. Draft warm messages in the user's own voice. Use when the user wants to clean up, grow, or refocus their network.

Build and fix Nuxt 4 apps with safe SSR, matching hydration, route rules, lazy loading, and data fetching with useFetch and useAsyncData. Use for page data, dynamic routes, browser-only code, caching, and speed work.

Prepare a private project for a safe public release. Copy it to a clean staging folder, remove secrets and private data, check the result, and add release files. Use for '/opensource', 'open source this', 'make this public', 'prepare for open source', or checks before a public GitHub release.

Build useful monitoring dashboards for Grafana, SigNoz, and similar tools. Use this skill to turn metrics into clear views that help operators find problems and take action.

Build and run a team of up to five agents in parallel

Plan and run coding work with many Claude Code agents through Claude DevFleet. Each agent works in its own Git worktree. Review the plan, track work, and read clear reports.

Patient safety checks for health app releases. Tests CDSS rules, PHI leaks, data safety, care steps, and system links. Blocks release when a critical check fails.

Build or refactor apps with clear ports and adapters. Use for domain rules, use cases, HTTP or CLI entry points, databases, queues, third-party APIs, tests, and code that is tied to a framework.

Turn a one-line goal into a clear build plan for a large coding project. Use it for work that needs many pull requests, sessions, or agents. Each step has enough context for a new agent to start with no prior chat. The plan includes dependencies, parallel work, review gates, common risks, rollback steps, and a safe way to change the plan. Trigger when the user asks for a blueprint, roadmap, or plan for complex work. Do not trigger for work that fits in one pull request, needs fewer than three tool calls, or when the user says to just do the work.

Search PubMed and NCBI E-utilities for health papers, MeSH terms, PMIDs, abstracts, citations, and saved search logs.

Safe Redis patterns for caching, sessions, rate limits, locks, messaging, key design, and connection setup in production apps.

Choose regex or an LLM to parse structured text. Start with regex. Use an LLM only for hard cases with low trust.

Build and edit Remotion videos with React. Use this skill for animation, audio, captions, charts, 3D scenes, media files, timing, trimming, and scene changes.

Evidence-first repo work for ECC. Use when a user asks to run commands, check a repo, debug a CI failure, make a small fix, verify it, commit it, or push it.

Use a clear rubric to review papers, proposals, literature reviews, methods, evidence, citations, and other research work.

Design clear REST APIs with good names, status codes, page rules, filters, errors, versions, access checks, and rate limits.

Split a large RFC into small work units, run them in dependency order, check quality, and merge them safely.

Scan skills, find shared ideas, and turn them into clear rules. Add to, fix, or create rule files only after user approval.

Safe patterns for EMR and EHR apps. Covers patient visits, notes, prescriptions, clinical alerts, audit logs, and easy-to-use forms.

Safe error handling for TypeScript, Python, and Go. Use it to add typed errors, API error replies, retries, circuit breakers, logs, and clear user messages.

Write, review, or refactor safe and clear C++17, C++20, and C++23 code using the C++ Core Guidelines.

Install the external repo-scan skill from a fixed commit that can be reviewed first. Use this pointer before a source code audit when repo-scan is not installed. It supports audits across C++, Android, iOS, and web code. This pointer installs the skill. It does not scan a repo.

Prevent hidden token decimal bugs across EVM chains. Read decimals at run time, cache by chain and token, handle bridged tokens, and safely scale values.

Build an automated agent that collects public data, adds useful labels or summaries, and saves the results. Use it for jobs, prices, news, GitHub projects, sports, events, and other public sources. It can run on a set schedule with GitHub Actions and save data to Notion, Google Sheets, or Supabase.

Turn folders of UI screenshots or design exports into Vue 3 components. Supports Vant, Element Plus, and Ant Design Vue.

Secure Perl code with taint mode, strict input checks, safe file and process use, DBI placeholders, web guards, and perlcritic checks.

Audit, plan, and make SEO fixes for technical SEO, page content, structured data, Core Web Vitals, internal links, sitemaps, robots.txt, and keyword mapping. Use when a user wants better search reach, an SEO audit, clear SEO fixes, schema markup, or a content plan.

Audit Claude skills and commands. Supports a quick scan of changed skills and a full review in small batches.

Search local skills, marketplaces, GitHub, and the web before you create, build, find, fork, or extend a Claude Code skill.

Measure speed, save a baseline, find slowdowns after a code change, and compare stack choices.

Verify a Spring Boot project before a pull request, large change, or release. Check the build, code rules, tests, coverage, known security flaws, format, and Git diff.

Review or add security in Java Spring Boot apps. Use for login, access rules, input checks, CSRF, secrets, headers, rate limits, file uploads, logs, and library risks.

Use Swift 6.2 approachable concurrency. Keep code on the caller's actor by default. Use @concurrent for clear background work. Use isolated conformances for MainActor types.

SwiftUI patterns for state, view layout, safe navigation, speed, previews, and modern iOS and macOS apps.

Use this skill to build or fix tinystruct app modules and small services. Covers routes, Context, Builder JSON, settings, tests, and actions that work in both CLI and HTTP modes.

Record polished web app demo videos with Playwright. Use when the user asks for a UI demo, walkthrough, screen recording, product tour, or tutorial. Explore the app, test the full flow, then record a WebM video with a visible cursor and clear timing.

Trace each user button, form, and touch point through its full state change path. Find bugs where calls work alone but undo each other, race, or leave the UI in the wrong state. Use this when a button looks broken, normal checks find no cause, or a large change touches shared state.

Find official USPTO patent and trademark records, check case status and ownership records, and keep a clear research log that others can repeat.

View, understand, edit, and search video or audio with VideoDB. Use local files, public URLs, YouTube links, RTSP streams, or live desktop capture. Create stream links, transcripts, scene indexes, clips, subtitles, overlays, audio tracks, alerts, and new media. Also transcode media and change its size, frame rate, codec, quality, or shape.

Translate visa document images into English and make a two-page PDF with the source image and translation.

Rank warm intro paths on X and LinkedIn. Use it to score bridges, find network gaps, and choose between a warm intro and a cold message. Use it only when the user needs the ranking engine, not a full lead or network care plan.

Create PNG image sets for Xcode asset catalogs from SF Symbols or Iconify. Use it to find icons, match an iOS or macOS project style, replace test icons, or make 1x, 2x, and 3x assets.

Plan safe deploys for Kubernetes, Docker, Vercel, and cloud hosts. Covers rolling, blue-green, and canary deploys, health checks, rollback, scaling, and updates with no downtime.