api-ts-dev
01Agent Claude Code
TypeScript API specialist — Hono routes, Effect-TS services, Drizzle ORM schemas, A2A protocol. Use for any work in api-ts/.
Agent Claude Code
TypeScript API specialist — Hono routes, Effect-TS services, Drizzle ORM schemas, A2A protocol. Use for any work in api-ts/.
Agent Claude Code
Art direction and visual quality gate for generative/NFT work and any pixel-facing surface. Judges rendered output, not markup — rasterizes, looks, and critiques against the long-form generative-art standard. Use before shipping any collection, card renderer, or visual system.
Agent Claude Code
Python Telegram bot specialist — handlers, services, models, swap logic, wallet management, WhatsApp, copy trading, perps, token security. Use for any work in bot/, api/, database/, or tests/.
Agent Claude Code
Brand system alignment across every Suwappu surface — showcase, webapp, mobile, bot copy, NFT renderers, docs. Enforces single-sourced design tokens and a consistent voice. Use when a surface may have drifted, or before shipping anything customer-facing.
Agent Claude Code
Blockchain integration specialist — add new chain support end-to-end across bot, API, webapp, and config. Use when adding a new blockchain to Suwappu.
Agent Claude Code
Database migration specialist — add columns, tables, indexes to both Python SQLAlchemy models and TypeScript Drizzle schemas. Use for any database schema changes.
Agent Claude Code
Railway deployment and operations agent — Railway service deploys, health checks, log tailing, infrastructure troubleshooting. Use for deployment and ops tasks.
Agent Claude Code
Positioning, launch narrative, mint/landing copy, and distribution for Suwappu launches. Writes the story a collection or feature is sold on, grounded in what the product actually does. Use for mint pages, announcements, pricing copy, and campaign planning.
Agent Claude Code
Production incident responder — diagnose outages, check health endpoints, read CloudWatch logs, SSM into EC2, restart services, trace errors. Use when something is broken in production.
Agent Claude Code
Adversarial Opus reviewer for money-path diffs — swap execution, wallet/key handling, encryption/KMS, billing/Stripe/x402, fee math, seasons/points accounting, withdrawals, redemptions. Read-only. Use after ANY change a builder tagged MONEY-PATH, or before merging anything that moves funds, keys, or balances.
Agent Claude Code
Read-only research specialist — web + codebase research, competitor scans, tokenomics/economics, design critiques, best-practice surveys, market/UX investigations. Returns a tight, cited summary. Use instead of general-purpose for ANY research/audit/triage question (general-purpose is an untiered Opus catch-all …
Agent Claude Code
Post-implementation code reviewer — reviews recent changes for quality, security, performance, and patterns compliance. Read-only. Use proactively after any significant code change.
Agent Claude Code
Fast read-only reconnaissance — grep/glob sweeps, file-finding, registration audits, parse/boot-import gates, dead-button audits, repo triage. Returns conclusions, not file dumps. Use for any "where is X / does Y exist / audit all Z" question instead of grinding on the main loop.
Agent Claude Code
SDK and package developer — maintains packages/sdk (TypeScript), packages/sdk-python, and packages/openclaw. Keeps SDKs in sync with API changes. Use for SDK work or when API changes affect client libraries.
Agent Claude Code
Security auditor for DeFi bot — OWASP top 10, wallet encryption, token security, input validation, dependency scanning, secret detection. Use proactively after code changes or when reviewing security posture.
Agent Claude Code
Next.js showcase site specialist — showcase/ homepage, pricing, contact/enterprise lead forms, marketing pages, premium visual polish, design tokens. Use for any work in showcase/.
Agent Claude Code
Heavy cross-service planner (Opus) — use ONLY for genuinely large, multi-service architecture/planning that wants a second Opus brain. Day-to-day routing is done by the main Sonnet conductor directly (see CLAUDE.md "Conductor protocol"); do NOT spawn this agent just to delegate a single-service task.
Agent Claude Code
Cross-chain swap debugger — trace failed transactions, diagnose quote errors, debug bridge issues, analyze token security problems, debug RPC/balance issues. Use when investigating swap/transaction/balance failures.
Agent Claude Code
Test engineering specialist — write pytest and vitest tests, validate coverage, run regression suites, generate test fixtures. Use when adding tests or verifying code changes.
Agent Claude Code
React webapp specialist — Telegram Mini App components, hooks, contexts, pages. Use for any work in webapp/ or packages/shared/.
Command Claude Code
Parallel attacker-minded audit: fan out one security-auditor per attack surface, each streaming findings to disk, then dedupe/rank and file issues. Usage: /audit-fleet [scope hint].
Command Claude Code
Attacker-minded security audit of scoped files. Traces auth + money-path data flows, streams compact findings incrementally (never one giant end-of-session JSON dump). Usage: /audit [attack-class].
Command Claude Code
Treat one confirmed bug as a class: reproduce, fix, generalize the class, sweep the repo, and fix every instance with one commit each. Usage: /bugclass.
Command Claude Code
Deploy Suwappu services on Railway. Usage: /deploy [prod|dev] [python-api|python-worker|terminal|api-ts|showcase|all].