kun

Multi-slide bilingual brand carousels — Claude writes the deck, kun renders Anthropic-styled slides at exact platform sizes, a human approves, channels receive

Draft, stage, and publish brand social posts — Claude drafts, /higgs renders, a human approves, Hermes relays

Convert a file or URL to Markdown via MarkItDown (PDF, Office, images, audio, web)

Full pipeline — idea to production (chains every stage)

Autonomous block QA — detect, adversarially verify, fix safe tiers, hand the residual to a human

Technical spec — data model, file plan, refined acceptance criteria

Self-update — read the latest releases across the engine's whole surface (Anthropic, stack, services, practice), diff against the engine, adopt or propose

Run the captain's weekly Monday-plan / Wednesday-check / Friday-review cycle

Check Claude Code config health across the team + engine drift

Generate and edit photos and videos for databayt org marketing, ads, and prompts using Higgsfield AI

Autonomously authenticate to any web service or CLI on Abdout's behalf using the macOS-Keychain-stored databayt identities, then reach protected pages and fetch original assets behind the login wall. Use whenever a page redirects to a login wall, a CLI reports "not authenticated", the user says "log into X / sign me into X / auth", or a clone/asset-fetch task needs content behind a paywall or dashboard. Minimizes interaction — only pauses for a one-time code when no non-interactive path exists.

One spell — handover, check, ship, watch, then comment + close the issue

Analyze a repo's patterns and generate a tailored Claude Code config

Clone a pattern/component from a source repo OR mirror a live URL section pixel-exact

Plan and review the content calendar

API spend breakdown and cost-optimization analysis

Crawl and snapshot Anthropic docs/assets

Manage databayt credentials — API keys via Keychain, web logins via Safari

Incident response workflow for production issues

Extract organizational patterns and conventions from repos + history

Check all deployments, costs, and uptime across databayt products

Cross-repo dependency audit and upgrade coordination

Look up a canonical cross-repo pattern (form, table, auth, ...)

Calculate and compare pricing tiers for a databayt product

Load a selective config profile by role (engineer/business/content/ops)

Generate a client proposal via the revenue agent

Quality gate — typecheck + build + visual verify before shipping

Logic layer — server actions with auth, validation, tenant isolation

UI verification — runs niche quality keywords on a URL or block

Capture a feature idea as a structured GitHub issue

Create a GitHub issue in the appropriate databayt repo

Architecture plan — tech decisions from an approved spec, before code

Auto-fix a user-reported issue end to end

Data layer — Prisma model, migration, and Zod validation

Promote a checked build to production (Vercel --prod)

Task breakdown — dependency-ordered work list from an approved plan

Post-deploy verification — visual + console + smoke test

UI layer — page (mirror pattern), content, form, table columns, i18n

Surface the right CEO book + operating move for a business or leadership decision — hiring, firing, pricing, positioning, what to build, customer development, fundraising, prioritization, runway, strategy. Pull in automatically when a decision is on the table; no slash command needed. /canon also lists all 10.

shadcn/ui knowledge pack — CLI, MCP, registry, skills, directory, and the docs-block pattern. Say "shadcn" to load it; "shadcn docs" = the docs-block style.