Explore

Audit one section of the docs content against the code, and report what is wrong, stale, fragile, off-voice, or missing.

Guide for making database schema changes in Platypus using Drizzle ORM — editing the schema, pushing to a dev database, and generating the migration that ships.

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

Skill for integrating Better Auth - the comprehensive TypeScript authentication framework.

Ask which skill or flow fits your situation. A router over the skills in this repo.

Response-body conventions for the Platypus backend API — the `error` key for failures, the `message` key for 2xx status messages, and when to throw a typed error instead of returning a response. Use when adding or changing a backend route or its tests.

Guide for adding new API endpoints to the Platypus backend using Hono.js — routing, tenant scoping middleware, validation, and database access.

Turn any workflow, conversation, or demonstrated process into a reusable SKILL.md. Use when the user says "turn this into a skill", "make a skill from this", "save this workflow", "remember how to do this", or after completing a complex multi-step task that should be repeatable. Also triggers on "tanítsd meg magad", "jegyezd meg ezt a folyamatot", "csinálj ebből skill-t".

List, inspect, patch, or delete skills from ~/.claude/skills/. Use when the user asks about available skills, wants to modify an existing skill, or when a retrospective proposes skill changes. Trigger on "/skills" command or skill-related retrospective actions.

Analyze the current session for improvement opportunities in skills, memory, and workflow. Spawns a sub-agent for unbiased analysis. Use when a session involved complex problem-solving, error recovery, user corrections, or multi-step workflows. Trigger on "/retrospective" command or at session end after significant work.

Merge a stack of GitHub PRs sequentially when they share files and will cause cascading conflicts. Triggers when user says "merge the PRs sorban" or similar, and the PRs come from external forks (cannot push back to PR branch).

A Marveen-flotta minden kolléga-asszisztensére érvényes, owner-független biztonsági és adatkezelési higiénia (Drive-írás hatóköre, login-automatizálás/credential-escalation, más megbízó adatainak védelme). Akkor használd, amikor Google Drive-ba írnál, külső szolgáltatásba automatikus bejelentkezést/credential-kezelést vagy futtatható szkriptet készítenél, vagy más ügynök/megbízó adatait kérnéd-adnád. A telepítés-specifikus (owner) kiegészítések a saját CLAUDE.md "Flotta-szabályok" blokkjában vannak.

Marveen flottában új channel plugin (slack-channel, telegram, stb.) telepítésekor a user-szintű ~/.claude/settings.json enabledPlugins minden agent-nek loadolja a plugin server-t, és ha az egy Socket Mode connection-t nyit (Slack), akkor TÖBB agent egyszerre nyit ugyanazon a workspace-en kapcsolatot. Akkor használd, ha egy agent Socket Mode-os channel plugin-t használ és inbound event-ek "fele eltűnik".

A Marveen Bridge és a "Szolgáltatás-lapok" ajánlása és elmagyarázása a gazdának. Három trigger. (1) PORT-ESEMÉNY, amikor az ügynök lokális porton futó szolgáltatást indít vagy telepít a Marveen gépén (CRM, Postiz, dev-szerver, admin felület). (2) Telepítés utáni korai beszélgetés, egyszer. (3) KÉRDÉS, amikor a gazda azt kérdezi, hogyan érheti el kényelmesebben a rendszert, mit lát a saját gépén vagy telefonján, hogyan lehet gördülékenyebb a használat, van-e asztali app vagy távoli hozzáférés.

A fő-ágens eljárása, amikor egy sub-ágens jóváhagyást kér az approval API-n keresztül ([APPROVAL_REQUEST] inter-agent üzenet). Kiküldi a kérést a tulajdonosnak, feldolgozza a szöveges válaszát, és lezárja az approvalt. Akkor használd, ha [APPROVAL_REQUEST] kezdetű inter-agent üzenetet kapsz.

AI fleet project execution (orchestrator=PM, marketing agent, backend dev agent, video agent). Fast-iteration architecture pivots and inter-agent task delegation across multi-hour sessions. Use when a user assigns a multi-agent project with "take it as a team" instruction.

On-demand diagnózis egy sikertelen frissítés + auto-rollback után. A dashboard "Frissítés diagnózisa" gombja indítja (opt-in). NEM cron-vezérelt.

Esti napzáró. A napi naplóba beírja, ami az utolsó bejegyzés óta történt (új emlékek, mozgott kártyák), hogy a délután és az este ne maradjon ki a naplóból.

Proaktív hírszerző -- napi brief 07:00-kor az intel registry-ből (store/intel.db). Az intel-collector párja, azzal együtt kapcsold be.

Proaktív hírszerző -- óránkénti gyűjtés a saját figyelt témáidban, eredmény az intel registry-be (store/intel.db). A napi brief (intel-napi-brief) ebből épül.

Minden körben átnézi az ELŐZŐ KÖR ÓTA történteket, menti a fontosat, és skill-eket generál ha volt komplex munka

2 percenként ellenőrzi, maradt-e megválaszolatlan bejövő üzenet a beszélgetés-ledgerben, és ha igen, felszínre hozza, hogy a futó session válaszoljon rá respawn nélkül

Éjszakai analízis-loop az aznapi memóriákról, naplóról és kanban-állapotról. Generál 4 priorizált akció-javaslatot reggelre.

Generate one contrarian, non-obvious insight on whatever the user is thinking about. The thing nobody else would say.

Generate 3 reply options + a recommendation for any incoming message thread. The user picks; /draft sends.

Generate a prioritized punch list from open commitments + tasks, right now. Get Shit Done.

Drafts outbound messages (WhatsApp, SMS, email) in the user's voice. Studies recent messages in a thread to mirror cadence. Never sends without explicit approval.

Design, review, or implement AgentOS UI and UX work using the project’s operator-console visual system. Use for React/Tailwind pages, cards, dialogs, mobile layouts, theme work, navigation, forms, empty/loading/error states, and UI consistency reviews in AgentOS.

Use AgentOS as the operator control plane above OpenClaw for starting, inspecting, and diagnosing a digital workforce.

Mine coding agent logs (Codex/Cursor/session histories and similar telemetry) to discover high-value candidate skills, then draft structured skill creation/reuse recommendations.

Hunt existing integration patterns and design vendor/framework-agnostic API, RFC, SDK, and integration plans for new external vendor integrations.

Remove AI-style noise from staged and branch diffs while preserving behavior and repository style.

Optimize LLM prompts, tools, and agents in Opik using standardized optimizer workflows (prompt optimization, tool optimization, and parameter tuning), dataset/metric wiring, and result interpretation.

Audit, compare, and document Opik integrations across Python SDK, TypeScript SDK, and OTEL/API backend. Use when adding a new integration, reviewing an existing one, or generating first-pass integration docs and gap analysis.

Create managed Git worktrees from a verified healthy owner, verify dependency reuse, and report explicit checkout and recovery outcomes.

Qualify and promote a pinned Bun fork build for a Linux OpenClaw service, including upstream sync, release proof, native workload admission, and rollback planning.

Maintain, verify, and release graincrawl, the local-first Granola archive CLI, including SQLite archive behavior, read-only Granola source boundaries, Homebrew tap packaging, and crawlkit-powered TUI/snapshot surfaces.

Inspect or switch the local Git and GitHub download guard using an already-installed compatible low-data command. Use for airplane mode, flight or metered connections, returning online, or requests to enable, disable, or automatically select low-data protection on this machine.

Use when inspecting a ghcrawl SQLite store, pulling GitHub issue/PR data, refreshing summaries, embeddings, and clusters, or extracting one cluster and its evidence through the ghcrawl CLI.

Maintain and release the crawlkit Go library, preserving downstream compatibility for gitcrawl, slacrawl, discrawl, and notcrawl.

Interpret Crabpot and OpenClaw performance dashboard metrics, import-loop profiles, runtime profiles, and branch-to-branch perf deltas without over-reading noisy samples.

Recover Codex or Claude sessions from tmux panes, saved cockpit snapshots, session history, and recent agent logs without overwriting useful restore evidence. Use when a pane crashes, detaches, loses visible context, needs a resume command, or requires a safe targeted restore plan.

Prepare and check trusted macOS hosts for physical Android and iPhone tests with agent-device, ADB, and Xcode, or set up iPhone Mirroring separately.

Search local documents, files, notes, and knowledge bases. Index directories, search with BM25/vector/hybrid, get AI answers with citations. Use when user wants to search files, find documents, query notes, look up information in local folders, index a directory, set up document search, build a knowledge base, needs RAG/semantic search, wants to start a local web UI for their docs, or asks what was said in past coding-agent sessions.

Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to "dogfood", "QA", "exploratory test", "find issues", "bug hunt", "test this app/site/platform", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.

Update the third-party JavaScript vendored into the app for the artifact, diagram, and math WebViews (KaTeX, mermaid, marked, highlight.js, Tailwind, Babel, React). Checks what is outdated, reads each pin's rationale before proposing a bump, re-downloads from the official source, verifies the lock, and runs the gates. Ends by naming what must be checked on a device, because a broken WebView asset fails silently. Use when bumping a vendored web library, adding a new one, or investigating whether a rendering bug comes from a stale pin.

Sync the Switchboard client with a newer official LibreChat server version — a stable release, a release candidate, or a PARTIAL sync up to an untagged upstream commit (e.g. current dev HEAD). If the client is ALREADY at the target commit, runs an AUDIT of the last synced delta to catch anything that was missed. Coordinates two dynamic workflows (discovery, implementation) around human decision gates.

Audit localization / i18n coverage across the compose-resources surface (10 modules x 9 locales). Finds strings that exist in English but not in some or all languages, keys left as untranslated English stubs inside a translated file, stale keys the base dropped, and English literals that never reached a strings.xml at all. Use when asking "which features shipped English-only?", after landing a feature that added user-facing strings, before a release that claims multi-language support, or when a translator asks what still needs doing. Audit-only — produces a report, never writes translations.

Audit open dependabot PRs in this repo. Spawns one investigator per PR (each with a codebase-walk sub-agent), produces a comparison table with per-PR risk + recommendation, and proposes a merge order that minimizes rebase churn. Audit-only — stops at recommendations; user authorizes execution. Use when dependabot has stacked up multiple open PRs and you want a structured pass before merging.

Daofy for Delphi 用法 skill（按需加载）。当实际需要调用 delphi_file 写入/修改 Delphi 文件、 需要做编译/审计/自动化测试，或遇到 Trae 等只暴露 run_mcp 的客户端包装时使用。 强制路由规则(Delphi 文件必用 delphi_file)由 Daofy Rule 与 initialize.instructions 常驻保证，本 skill 只讲"怎么用"。

Run WhatsApp marketing campaigns with ReachOwl: import a phone list or target a WhatsApp group, write personalized message sequences with variables and follow-ups, launch, track replies, and retry failed messages. Use for WhatsApp bulk messaging, WhatsApp broadcasts to customer lists, WhatsApp group outreach, or personalized WhatsApp follow-ups.

ReachOwl social media marketing automation for Facebook, Instagram, and WhatsApp: social listening, signal-based prospecting, cold DM campaigns, WhatsApp marketing, Facebook group posting, comment-to-DM automations, Instagram growth, lead finding, and CRM follow-ups through the ReachOwl MCP server. Use when the user mentions ReachOwl or wants to automate Facebook, Instagram, or WhatsApp marketing and doesn't name a specific ReachOwl feature.

Social listening with ReachOwl keyword monitors: track keywords across Facebook groups, Instagram, Reddit (subreddits), and Quora, and review the posts found. Use when the user wants brand monitoring, keyword alerts, to find people asking for a product or service, to discover relevant discussions, or to see posts a monitor found.

Signal-based prospecting with ReachOwl: find people who just posted a need ('looking for a realtor', 'can anyone recommend a CRM') in Facebook groups, Instagram, Reddit, or Quora, qualify them, and reach out with a personalized DM or reply. Use when the user wants intent-based leads, buyer-intent monitoring, to turn social listening into outreach, or to DM people who commented on or reacted to a relevant post.

Manage the ReachOwl sales pipeline (CRM stages) and reusable message templates for follow-ups. Use when the user wants to add, rename, reorder, or delete pipeline stages, or write, save, or update follow-up message templates tied to a stage.

Send ReachOwl events to other apps with webhooks: new keyword-monitor posts, sent messages, replies, accepted friend requests, and failures. Use when the user wants notifications in Slack, a CRM, Zapier, Make, n8n, or their own server when something happens in ReachOwl, or wants to list, change, or remove webhooks.

See which Facebook, Instagram, and WhatsApp accounts and browsers are connected to ReachOwl, pick the account to run campaigns from, and manage team members (invite, change role, remove). Use when the user asks which accounts they can use, why an account isn't available, or wants to add a teammate to ReachOwl.

Look up and retry ReachOwl message queue items for a specific lead in a campaign. Use when a DM, friend request, or follow-up failed or is stuck, when the user asks whether a message was sent to someone, or wants to resend or reschedule a step.

Connect and verify a ReachOwl account in the ReachOwl MCP server using the user's ReachOwl API token. Use when setting up ReachOwl, when any ReachOwl tool reports a missing or invalid token, when the user pastes a ReachOwl token, wants to log in with email and password, switch accounts, or log out.

Make a raw authenticated call to the ReachOwl V1 REST API through the ReachOwl MCP server when no named ReachOwl tool covers the request, such as a query parameter or field the named tools don't expose. Use only after checking the other ReachOwl skills.