joelhooks

Run a comprehensive health check of the joelclaw system — k8s cluster, worker, Inngest, Redis, Typesense/OTEL, tests, TypeScript, repo sync, memory pipeline, pi-tools, git config, active loops, disk, stale tests. Outputs a 1-10 score with per-component breakdown. Use when: 'system health', 'health check', 'is everything working', 'system status', 'how's the system', 'check everything', or at session start to orient.

Operate the joelclaw Kubernetes cluster — Talos Linux on Colima (Mac Mini). Deploy services, check health, debug pods, recover from restarts, add ports, manage Helm releases, inspect logs, fix networking. Triggers on: 'kubectl', 'pods', 'deploy to k8s', 'cluster health', 'restart pod', 'helm install', 'talosctl', 'colima', 'nodeport', 'flannel', 'port mapping', 'k8s down', 'cluster not working', 'add a port', 'PVC', 'storage', any k8s/Talos/Colima infrastructure task. Also triggers on service-specific deploy: 'deploy redis', 'redeploy inngest', 'livekit helm', 'pds not responding'.

Deploy the system-bus-worker to the joelclaw Kubernetes cluster from local machine. Use when syncing changes in packages/system-bus to k8s, especially because the GitHub Actions deploy job targets a non-existent self-hosted runner and cannot complete deploys automatically.

Operate Talon, the Rust infrastructure watchdog daemon that supervises the system-bus worker and monitors k8s. ADR-0159.

Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions.

Search captured agent Runs and raw local/remote Pi sessions, especially dark-wizard sessions, using the joelclaw sessions bridge. Use when the user asks to search sessions, find prior dark-wizard/Panda/pi/codex/claude context, recover conversation history, verify session indexing, or bypass stale rag_search_sessions/Typesense results.

Create new joelclaw skills with the idiomatic process — repo-canonical, symlinked, git-tracked, slogged. Triggers on 'add a skill', 'create skill', 'new skill', 'canonical skill', 'make a skill for', or any request to formalize a process or domain into a reusable skill.

Vercel AI Gateway expert guidance. Use when configuring model routing, provider failover, cost tracking, or managing multiple AI providers through a unified API.

Send messages, system events, and agent notifications through the openclaw CLI. Use when an agent needs to notify a user (Telegram, Slack, Discord), trigger another agent, broadcast updates, send system events, or coordinate agent-to-agent communication via swarmmail. Covers all three messaging layers - external channels, agent invocation, and internal swarmmail.

Handles version bumps and npm releases for the swarm-tools monorepo (opencode-swarm-plugin, claude-code-swarm-plugin, swarm-mail, swarm-queue). Use when: creating changesets, bumping versions, preparing releases, checking release status, debugging publish failures, verifying npm packages, or merging release PRs. Triggers: "release", "publish", "changeset", "bump version", "ship it", "new version", "create a release", "check npm", "verify publish", "/release"

Always-on rule-oriented guidance for claude-plugin agents. Use to align behavior, tool usage, and model-specific defaults while avoiding deprecated bd/cass references. Related skills: swarm-coordination, testing-patterns.

Multi-agent coordination patterns for OpenCode swarm workflows. Use when work benefits from parallelization or coordination. Covers: decomposition, worker spawning, file reservations, progress tracking, and review loops.

Always-on rule-oriented guidance for claude-plugin agents. Use to align behavior, tool usage, and model-specific defaults while avoiding deprecated bd/cass references. Related skills: swarm-coordination, testing-patterns.

Ralph loop pattern - Claude supervises while Codex (gpt-5.3-codex) executes implementation work. Use for autonomous coding loops with fresh context per iteration, validation gates, and git-backed persistence. Tools: ralph_init, ralph_story, ralph_iterate, ralph_loop, ralph_status, ralph_cancel, ralph_review.

Swarm CLI commands for workers - hivemind memory, hive tasks, swarmmail coordination. Use when working in a swarm context. Covers: swarm memory (find/store/get/stats), swarm cells (query/create/update/close), and coordination commands.

Multi-agent coordination patterns for OpenCode swarm workflows. Use when work benefits from parallelization or coordination. Covers: decomposition, worker spawning, file reservations, progress tracking, and review loops.

Build and maintain the native iOS client inside this turborepo (XcodeGen, SwiftUI, WebSocket feed). Use when creating/updating the iOS app, wiring turbo tasks, or documenting iOS workflows.

Publish the native iOS app to TestFlight from this repo. Use when archiving/exporting/uploading builds, bumping versions, checking App Store Connect processing status, or fixing common upload failures.

Deploy the atproto-agent-network project. Two separate Cloudflare Workers must be deployed independently — the network worker (API + agents) and the dashboard worker (highswarm.com). Use when deploying changes, fixing the dashboard, or updating agent configs.

Cloudflare Durable Objects patterns for agent state. Use when implementing agent DOs, WebSocket handling, hibernation, storage API, alarms, or DO-to-DO communication. Triggers on Durable Object, DO state, WebSocket server, hibernation, agent persistence.

Deploy applications and infrastructure to Cloudflare using Workers, Pages, and related platform services. Use when the user asks to deploy, host, publish, or set up a project on Cloudflare.

Pi agent runtime integration for Cloudflare. Use when implementing the agent loop, tools, extensions, session trees, or multi-provider LLM calls. Triggers on Pi, agent runtime, tool calling, streaming, session tree, extensions, self-extending agent.

D1 database patterns for encrypted agent records. Use when implementing schema, storing/querying encrypted records, migrations, or working with the records table. Triggers on D1, database, schema, records table, SQL, encrypted storage.

Envelope encryption patterns with X25519 for agent memories. Use when implementing key generation, DEK management, encrypting/decrypting records, key rotation, or sharing encrypted data between agents. Triggers on encryption, X25519, DEK, envelope encryption, key exchange, crypto.

Install, verify, and debug JoelClaw Codex session capture hooks and compare Codex/Claude/Pi capture health.

Search Joel's Pi, Claude, and Codex session history using the pi-tools session-reader flow. Use when the user asks to search past sessions, recover context, find prior decisions, inspect transcript history, continue old work, find support patterns from previous runs, or verify whether sessions are being captured by JoelClaw.

Resolve project-local issue tracker routing and publish Linear issues safely. Use before creating Linear issues, PRDs, phase breakdown issues, or when deciding whether Linear/GitHub/local docs is the configured tracker. Enforces capability ≠ routing: Linear requires project-local policy, team association, auth, and verified readback.

Create distinctive, crafted TUI components for pi using @mariozechner/pi-tui and @mariozechner/pi-coding-agent. Use when building interactive terminal UIs — custom components, overlays, dialogs, dashboards, widgets, data visualizations, animated elements, game-like interfaces, or any visual TUI work inside pi extensions or custom tools. Triggers on: 'build a TUI component', 'make a dashboard', 'create an overlay', 'interactive widget', 'terminal UI', 'custom component', 'pi-tui', or any request to create visual, interactive terminal interfaces. Also use when beautifying or redesigning existing TUI components.

Front inbox janitor for Joel's personal shitbox: Brain-mapped executive digests, contextual grouping, safe archive sweeps, and Front thread links. Use when asked to clean the shitbox, summarize inbox, run a digest, sweep personal Front, or tune inbox interest rules.

Post GitHub issue comments, PR reviews, and small file commits as the ShitRat GitHub App (`shitratgit[bot]`) instead of Joel. Use when reviewing PRs, commenting on issues, doing smoke tests, committing repo files through the GitHub API, or any workflow where GitHub authorship should be ShitRat.

Maintain, install, update, doctor, and release a ShitRat familiar across supported harnesses.

Maintain Cloudflare Agent Dungeon .brain notes as the durable project graph using PARA placement and source-backed receipts.

Query JoelClaw docs/PDF Brain for Old School Essentials rules while preserving Referee-only access and avoiding public corpus leakage.

Clone/update project-local source mirrors under .agent-sources, keep them out of commits, and read docs/source before making API claims.

Design and maintain composable data-backed Brain MDSvX components for pi-notes. Use when creating or editing `.brain/**/*.svx`, `.brain/components`, `.brain/data`, pi-notes Document Host components, review surfaces, or notes that should render JSONL, SQLite, Postgres, remote data, or reusable web shapes.

Create or operate local product-namespaced Pi review documents that can send feedback turns back to the spawning Pi session.

Create very clean, legible Excalidraw DAG/system diagrams for pi-notes. Use for architecture diagrams, feedback loops, data/control flow, state/lifecycle maps, or any request for a simple Excalidraw diagram where readability matters.

Portable credential management for AI agents using age encryption, session-scoped leases, auto-rotation, and killswitch. Use this skill when agents need secure, time-bounded access to API keys, tokens, or credentials without direct exposure to plaintext secrets.

Maintain Joel's macOS dotfiles as an agentic machine profile. Use when syncing configs from a live machine, adding Pi/Claude/Codex/ShitRat/JoelClaw setup docs, bootstrapping a new Joel-flavored machine, or reviewing dotfiles for secrets and local runtime leakage.

Write idiomatic Effect v4 TypeScript following official best practices from effect-solutions and the Effect source. Use when writing, reviewing, or refactoring Effect code: services (ServiceMap.Service), layers and dependency injection, error handling (Schema.TaggedErrorClass), data modeling (Schema.Class, branded types, variants), testing (@effect/vitest), HTTP clients (effect/unstable/http), CLI tools (effect/unstable/cli), config, observability, and project setup. Triggers on: 'Effect', 'effect-ts', '@effect/', 'Schema', 'ServiceMap', 'Layer', 'Effect.gen', 'Effect.fn', 'TaggedError', 'branded types', or any Effect-TS related code.

Control cmux terminal multiplexer from pi — read other terminal screens, send commands to other workspaces, manage panes/splits, set sidebar status/progress, send notifications, and open browser surfaces. Use when asked to: check what's running in another terminal, send a command to another pane, split the workspace, show progress in the sidebar, notify the user, open a URL in a browser pane, orchestrate multi-workspace workflows, or interact with cmux in any way. Triggers on: 'cmux', 'other terminal', 'other workspace', 'read screen', 'send command to', 'split pane', 'sidebar status', 'progress bar', 'notify', 'browser pane', 'terminal multiplexer'.

Manage Todoist tasks, projects, comments, reminders, and activity via the todoist-cli agent CLI. Use when: 'add a task', 'what's due today', 'check my inbox', 'complete a task', 'todoist', 'todo', 'task list', 'what did I finish', 'add a comment', 'set a reminder', 'daily review', 'search tasks', 'move task to project', 'what's overdue', or any task management request. All output is HATEOAS JSON with next_actions — parse result.tasks, result.comments, etc.