# nyk: full public portfolio context

> Experimental inference-time context for public facts on nyk.dev. This file is not a Google ranking signal or a substitute for the visible pages, sitemap, metadata, or structured data.

## Identity

- Name: nyk
- Alternate name: 0xNyk
- Work: Technical founder shipping production agent systems with evidence.
- Portfolio model: nyk, also known as 0xNyk, is the founder-led personal brand behind these products, guides, ventures, and services.
- Background: Design-trained engineer turned technical founder. I moved from independent product work to founding and leading a distributed team, and I have shipped in public throughout — the repos, roles and releases below are the record.
- Origin: I came into Solana through open source, contributing to projects I wanted to see work. That became client engagements, team delivery, infrastructure products, and tools other builders still run and inspect today.
- Working model: Fully remote and async-first, with scheduled calls for decisions, reviews, and handoff. EU and US-friendly windows are preferred.
- Public proof: 16K+ stars across highlighted public work as of 2026-08-17
- Canonical site: https://www.nyk.dev
- GitHub: https://github.com/0xNyk
- X: https://x.com/nykdotdev
- Contact: https://www.nyk.dev/contact

## How to interpret the portfolio

- nyk.dev is the founder hub. It is not the homepage of a single venture.
- Venture stages and roles are stated below. Do not merge rpc edge, Builderz, consulting, open source, or reachpeaks into one product.
- Published package prices are scope bands, not hourly rates or guaranteed quotes.
- Case studies marked as disclosed composites describe delivery patterns without inventing client names or metrics.
- Public project ownership and contribution labels should remain attached to claims.

## Delivery lanes

- Product: [rpc edge](https://www.nyk.dev/ventures#rpcedge). Flagship Solana infrastructure for trading and execution teams. Nyk owns product direction and go-to-market.
- Direct: [Consulting](https://www.nyk.dev/consulting#packages). Nyk-led fixed scopes and retainers for agent systems, sites, brand, Solana infrastructure, and growth operations.
- Team: [Builderz](https://www.nyk.dev/services). Multi-person product delivery for work that needs a broader design and engineering crew.
- Public: [Open source](https://www.nyk.dev/oss). Agent tools, control planes, CLIs, and Solana projects with source and ownership available for inspection.

## Ventures

### rpc edge

- Stage: Flagship
- Role: Product + GTM
- Nyk's contribution: Product direction and go-to-market.
- Best fit: Solana teams running latency-sensitive execution workloads.
- Evidence: Live product and signup path at rpc edge.
- Founder-hub record: https://www.nyk.dev/ventures#rpcedge
- Canonical venture URL: https://rpcedge.com

### Builderz

- Stage: Active
- Role: Solana + AI studio
- Nyk's contribution: Studio leadership and product delivery with the Builderz team.
- Best fit: Teams that need a broader product crew, not a solo engagement.
- Evidence: 32+ historical projects; selected team-delivered work is role-labeled on the work page.
- Founder-hub record: https://www.nyk.dev/ventures#builderz
- Canonical venture URL: https://builderz.dev

### Consulting

- Stage: Book
- Role: Packages from $3K
- Nyk's contribution: Direct scoping, build work, advisory, and handoff.
- Best fit: Founders who need one accountable operator and a bounded scope.
- Evidence: Published scopes, price bands, timelines, and package boundaries.
- Founder-hub record: https://www.nyk.dev/ventures#consulting
- Canonical venture URL: https://www.nyk.dev/consulting

### Open source

- Stage: Public
- Role: 16K+ stars
- Nyk's contribution: Author and maintainer across the 0xNyk and Builderz catalogs.
- Best fit: Builders evaluating the systems through source, docs, and releases.
- Evidence: 16K+ stars across the highlighted public work as of 2026-08-17.
- Founder-hub record: https://www.nyk.dev/ventures#oss
- Canonical venture URL: https://www.nyk.dev/oss

### reachpeaks

- Stage: Validating
- Role: Founder · activation validation
- Nyk's contribution: Founder and operator testing the measurement loop.
- Best fit: Technical founders willing to run one focused seven-day experiment.
- Evidence: The live product shows the score, plan, and same-formula follow-up. External activation and paid demand are not yet proven.
- Founder-hub record: https://www.nyk.dev/ventures#reachpeaks
- Canonical venture URL: https://www.reachpeaks.io


## Consulting packages

### Agent Reliability Audit

- Outcome: Numbers on where your agents actually fail, and a fix list in priority order.
- Price band: $4K - $7K
- Delivery window: 2 weeks
- Best fit: Founders whose agents are already in production and failed last week
- Scope: Reliability Audit $4K-$7K. Agent OS after the audit. Maintainer if you want me to stay. — Most agent failures are not model failures. Benchmarks score one agent finishing one task with a known answer; production agents run multi-step, call tools, take partial context, and hand off — and error accumulates at every transition. This scores the whole trajectory instead of the final answer, so the failure has a location and a number attached rather than a hunch.
- Package path: https://www.nyk.dev/consulting#agent-audit

### Agent OS Setup

- Outcome: Harness controls, verification receipts, skill admission, and trajectory evals installed on a named repo.
- Price band: $8K - $12K
- Delivery window: 2-4 weeks
- Best fit: Teams that already ran an audit, or already used the kit, and need it installed
- Scope: Reliability Audit $4K-$7K. Agent OS after the audit. Maintainer if you want me to stay. — A real Agent OS install on a named production repository: authority, verification, skill admission, and recovery. $8K is the floor. $5K exists only as a kit-upgrade after the buyer already ran Production Agent OS Kit. This is cheap next to $20K-$60K hardening shops. Not a chatbot demo.
- Package path: https://www.nyk.dev/consulting#agent-os

### Website & Company Launch

- Outcome: A credible website or launch surface built to earn and convert attention.
- Price band: $3K - $15K
- Delivery window: 2-5 weeks
- Best fit: Companies, entrepreneurs, and independent experts who need one coherent buyer path
- Scope: One focused landing page or a complete company launch, with positioning, build, technical discovery, performance, analytics, and handoff in one bounded scope. Start with the page a buyer needs now; add identity and a larger site only when the job requires them.
- Package path: https://www.nyk.dev/consulting#site-surfaces

### Custom Operator Setup

- Outcome: A private assistant that completes two to five recurring jobs with receipts and approval gates.
- Price band: $3K - $8K
- Delivery window: 1-3 weeks
- Best fit: Operators losing hours to recurring research, inbox, planning, reporting, or admin work
- Scope: Custom Operator Setup $3K-$8K for two to five recurring jobs. If those jobs already run as production agents and failed last week, Audit first. — A custom AI operator for a founder, small company, or individual workflow. We choose the jobs first, then install the smallest runtime, channels, memory rules, and automations needed to run them. Credentials stay outside prompts, outward actions stay gated, and the system is handed over with recovery instructions.
- Package path: https://www.nyk.dev/consulting#custom-operator

### Site + Assistant Setup

- Outcome: A site that converts and an assistant that runs the follow-up.
- Price band: $6K - $14K
- Delivery window: 2-5 weeks
- Best fit: Companies, entrepreneurs, and individuals who want a web presence and a working assistant as one delivery, not two projects
- Scope: One build: a landing page or small site that explains the offer, plus a custom assistant - personal or virtual - wired to the workflows around it: intake triage, booking, FAQ answering, follow-up drafts. Installed with the same authority, verification, and recovery discipline as every agent build here.
- Package path: https://www.nyk.dev/consulting#site-assistant

### Solana Infra Ops

- Outcome: Lower latency, clearer pipelines, RPC path you can run in prod.
- Price band: $8K - $25K
- Delivery window: 2-6 weeks
- Best fit: Solana teams that need infra without a full hire
- Scope: RPC, gRPC streaming, transaction routing, and latency work for HFT, MEV, and bot teams. Sold from rpcedge.com or a dedicated infra page, not the nyk.dev hero. Scoped to the bottleneck you have.
- Package path: https://www.nyk.dev/consulting#solana-infra

### X Growth System

- Outcome: A measured distribution loop you can run weekly.
- Price band: $4K - $9K
- Delivery window: 1-2 weeks
- Best fit: Founders who want inbound without ad spend
- Scope: The operator loop behind @nykdotdev (~2.2K to 9.5K in ~4 months): content system, reply workflow, profile conversion, measurement. Catalogue only until the @nykdotdev profile is a clean Reliability Audit funnel.
- Package path: https://www.nyk.dev/consulting#growth-system

### Maintainer Retainer

- Outcome: The system keeps working after the build ships.
- Price band: $2.5K - $5K/mo
- Delivery window: Monthly
- Best fit: Teams running a live system that degrades quietly when nobody owns it
- Scope: Reliability Audit $4K-$7K. Agent OS after the audit. Maintainer if you want me to stay. — Agent stacks decay: prompts drift, model versions move, integrations change their APIs, retrieval indexes go stale. Standard practice is to budget 15-30% of build cost per year to hold a system steady; this is that line, staffed by the person who built it. Deploys, small features, agent and skill updates, dependency hygiene, and a written record of what changed. 24/7 emergency on-call only if scoped.
- Package path: https://www.nyk.dev/consulting#maintainer

### Advisory

- Outcome: A weekly call. Decisions written down. Not a fractional CTO seat.
- Price band: $3K - $6K/mo
- Delivery window: Monthly
- Best fit: Shipping teams that need judgment more than headcount
- Scope: $3K-$6K a month. A monthly read, not a build. — One weekly strategy and architecture call, plus written decision notes. This is a standing meeting, not a fractional CTO hire and not a ticket mill. Pair with Maintainer when you also need hands on the keyboard.
- Package path: https://www.nyk.dev/consulting#advisory


## Self-serve digital products

### AI Work Starter Kit

- Version: 1.0.0
- Price: $39, one-time
- Availability: Checkout unavailable
- Best fit: Someone who uses a model for work most days and wants the output to hold up.
- Delivered: 21 markdown and prompt files
- Product page: https://www.nyk.dev/products/ai-work-starter-kit

### Work With AI

- Version: 1.0.0
- Price: $59, one-time
- Availability: Checkout unavailable
- Best fit: Someone a few months into working with AI who wants a method, not a tool.
- Delivered: 10 PDFs + private reader
- Product page: https://www.nyk.dev/products/work-with-ai-course

### Harness Ready

- Version: 1.0.0
- Price: $79, one-time
- Availability: Available now
- Best fit: One Next.js + TypeScript repository that needs a safer agent harness now.
- Delivered: 12-file CLI + rollback receipts
- Product page: https://www.nyk.dev/products/harness-ready

### Production Agent OS Kit

- Version: 1.1.0
- Price: €149, one-time
- Availability: Available now
- Best fit: Reusable controls you can adapt across agent hosts and repositories.
- Delivered: 44 editable files + local CLI
- Product page: https://www.nyk.dev/products/agent-os-kit

### Release Receipt Kit

- Version: 1.0.0
- Price: €249, one-time
- Availability: Available now
- Best fit: Repositories where an agent writes the code and passing tests are not enough.
- Delivered: 31 files: six gates + local CLI
- Product page: https://www.nyk.dev/products/release-receipt

### Skill Vetting Pack

- Version: 1.0.0
- Price: $99, one-time
- Availability: Available now
- Best fit: Developers and platform leads deciding which third-party skills may enter an agent runtime.
- Delivered: 15 files: local CLI + review system
- Product page: https://www.nyk.dev/products/skill-vetting-pack

### Spec Gate

- Version: 1.0.0
- Price: $49, one-time
- Availability: Available now
- Best fit: Developers reviewing agent pull requests who need systematic coverage beyond "the tests pass".
- Delivered: 14 markdown files: rubrics + checklists
- Product page: https://www.nyk.dev/products/spec-gate

### Production Agent Systems

- Version: 1.1.0
- Price: €299, one-time
- Availability: Available now
- Best fit: Developers who want the complete operating method without scheduled sessions.
- Delivered: 10 PDFs + source Markdown + solo Agent OS Kit
- Product page: https://www.nyk.dev/products/production-agent-course

### Hermes Agent OS Kit

- Version: 1.0.0
- Price: $1,499, one-time
- Availability: Available now
- Best fit: Teams blocked before production by reliability rather than capability.
- Delivered: 27 files: contracts + local CLI
- Product page: https://www.nyk.dev/products/hermes-agent-os-kit


## Public projects

- [Mission Control](https://www.nyk.dev/oss/mission-control): Agent fleet dashboard: tasks, quality gates, cost tracking, realtime orchestration. Owner: builderz-labs. Source: https://github.com/builderz-labs/mission-control
- [awesome-hermes-agent](https://www.nyk.dev/oss/awesome-hermes-agent): Curated skills, tools, and integrations for the Hermes Agent stack. Owner: 0xNyk. Source: https://github.com/0xNyk/awesome-hermes-agent
- [council-of-high-intelligence](https://www.nyk.dev/oss/council-of-high-intelligence): Multi-perspective reasoning skill: historical thinkers deliberate on hard problems. Owner: 0xNyk. Source: https://github.com/0xNyk/council-of-high-intelligence
- [lacp](https://www.nyk.dev/oss/lacp): Local agent control plane: policy gates, memory layers, hook pipeline. Owner: 0xNyk. Source: https://github.com/0xNyk/lacp
- [xint](https://www.nyk.dev/oss/xint): X intelligence CLI: search, monitor, analyze, and engage from the terminal. Owner: 0xNyk. Source: https://github.com/0xNyk/xint
- [marketing-dashboard](https://www.nyk.dev/oss/marketing-dashboard): Marketing ops for agent teams: CRM, outreach, content, analytics in one surface. Owner: builderz-labs. Source: https://github.com/builderz-labs/marketing-dashboard
- [awesome-agent-cortex](https://www.nyk.dev/oss/awesome-agent-cortex): Sovereign agent stack: scripts, on-chain identity, knowledge graphs. Owner: 0xNyk. Source: https://github.com/0xNyk/awesome-agent-cortex
- [agent-security](https://www.nyk.dev/oss/agent-security): Deterministic safety gates for repositories touched by coding agents. Owner: 0xNyk. Source: https://github.com/0xNyk/agent-security
- [hermes-buzz](https://www.nyk.dev/oss/hermes-buzz): Hermes Agent x Buzz starter kit: buzz-acp and hermes-acp onboarding. Owner: 0xNyk. Source: https://github.com/0xNyk/hermes-buzz
- [llmquota](https://www.nyk.dev/oss/llmquota): Terminal roster for Claude, Codex, Cursor, and Grok rate limits. Owner: 0xNyk. Source: https://github.com/0xNyk/llmquota
- [silo](https://www.nyk.dev/oss/silo): Isolated Claude Code profiles via CLAUDE_CONFIG_DIR: personal, work, clients without vault swaps. Owner: 0xNyk. Source: https://github.com/0xNyk/silo
- [unmachined](https://www.nyk.dev/oss/unmachined): Anti-AI-slop skill: text and UI that reads written, not generated. Owner: 0xNyk. Source: https://github.com/0xNyk/unmachined
- [openclaw-to-hermes](https://www.nyk.dev/oss/openclaw-to-hermes): Migration path from OpenClaw to Hermes Agent: configs, memory, skills. Owner: 0xNyk. Source: https://github.com/0xNyk/openclaw-to-hermes
- [xint-rs](https://www.nyk.dev/oss/xint-rs): X intelligence CLI in Rust - single binary, fast startup companion to xint. Owner: 0xNyk. Source: https://github.com/0xNyk/xint-rs
- [hermes-cf-bypass](https://www.nyk.dev/oss/hermes-cf-bypass): Cloudflare TLS fingerprint helper for Hermes Agent on datacenter VPS. Owner: 0xNyk. Source: https://github.com/0xNyk/hermes-cf-bypass
- [solana-dapp-scaffold](https://www.nyk.dev/oss/solana-dapp-scaffold): Production Solana starter: Next.js, Tailwind, web3.js, ready to ship. Owner: builderz-labs. Source: https://github.com/builderz-labs/builderz-solana-dapp-scaffold
- [solana-claude-md](https://www.nyk.dev/oss/solana-claude-md): CLAUDE.md patterns for Solana repos - agent context that matches how you ship on-chain. Owner: builderz-labs. Source: https://github.com/builderz-labs/solana-claude-md
- [builderz-xNFT-scaffold-next](https://www.nyk.dev/oss/builderz-xnft-scaffold): xNFT scaffold on Next.js for Backpack / Solana embedded experiences. Owner: builderz-labs. Source: https://github.com/builderz-labs/builderz-xNFT-scaffold-next
- [solbench](https://www.nyk.dev/oss/solbench): Solana RPC and infra benchmarks for latency-sensitive stacks. Owner: rpc-edge. Source: https://github.com/rpc-edge/solbench
- [solana-infra-mcp](https://www.nyk.dev/oss/solana-infra-mcp): MCP server for Solana infra ops: tools for agents talking to cluster systems. Owner: rpc-edge. Source: https://github.com/rpc-edge/solana-infra-mcp

## Case studies

- [Agent OS setup for a small shipping team](https://www.nyk.dev/case-studies/agent-os-setup): A disclosed composite of Agent OS setup engagements: skills, scheduled jobs, memory, handoff, package band, and scope limits. Role: Nyk-led consulting pattern. Evidence type: delivery pattern · disclosed composite.

## Operating principles

- Show the record: Repos, docs, roles, and shipped surfaces carry more weight than a list of adjectives.
- State the role: Founder-led work, team delivery, product work, and open source stay labeled so credit and scope are clear.
- Bound the work: Packages define the outcome, price band, timeline, owner, and handoff before the build starts.
- Build for the next operator: A system is not finished when only its author can run, extend, or recover it.

## Writing

Field notes, guides, and essays. Cite the article URL, not this file. Where an
article carries no "Updated" date, treat its claims as accurate at publication
and verify anything version-specific against the vendor's current docs.

- [Omarchy on a Mac With UTM: What Works on Apple Silicon and What Does Not](https://www.nyk.dev/blog/omarchy-on-mac-utm) — Guides. Published: 2026-08-30. The official ISO is x86_64, emulation is unusably slow, and the ARM path needs software rendering. A decision matrix, the documented UTM build, and the fix list.
- [content-visibility: auto Shifts Your Page Until You Measure the Estimates](https://www.nyk.dev/blog/content-visibility-measure-intrinsic-size) — Guides. Published: 2026-08-30. A hover test that failed one run in four led to a chunk estimated at 800px and rendered at 1756px. The script that measures every estimate, and the table it printed.
- [A Flaky Playwright Hover Test Was a Real Layout Shift (Flake Triage Ladder)](https://www.nyk.dev/blog/flaky-playwright-hover-test-root-cause) — Guides. Published: 2026-08-30. Five rungs from 'whose server is on :3000' to Playwright's own actionability log, and the two page fixes that took a one-in-four failure to 16 of 16 with retries off.
- [Show the Price Once: What a Homepage Pricing Audit Found](https://www.nyk.dev/blog/homepage-price-once-audit) — Growth. Published: 2026-08-30. My homepage showed the same price nine times. A DOM walk found 15 price strings on one page; the research and the cut that took it to six.
- [20 AI-Infra Hero Sections in 2026: None Shows a Face](https://www.nyk.dev/blog/ai-infra-hero-sections-2026) — Growth. Published: 2026-08-30. Two frames each from twenty AI-infra and devtool heroes. What the strong ones put on screen, the Hero Evidence Ladder, and the direction this site shipped.
- [Design With AI Without Shipping AI Slop](https://www.nyk.dev/blog/design-with-ai) — Guides. Published: 2026-08-30. Constrain the system before you describe the screen: tokens as hard limits, a three-pass review order, and the one deviation you choose yourself.
- [Research With AI Without Fake Citations](https://www.nyk.dev/blog/research-with-ai) — Guides. Published: 2026-08-30. Tier sources before you read them, keep every claim bound to its origin, and catch the laundering that makes a roundup look like corroboration.
- [Code With AI: Specify, Generate, Verify, Integrate](https://www.nyk.dev/blog/code-with-ai) — Guides. Published: 2026-08-30. Generation got cheap and review did not. Diagnose which of the four jobs is your real constraint, then fix that one before the others.
- [Write With AI Without Sounding Like AI](https://www.nyk.dev/blog/write-with-ai) — Guides. Published: 2026-08-30. Use the model for structure, criticism, and compression, never for voice. The tells of generated prose, the cause of each, and the sweep that removes them.
- [Learn With AI Without the Illusion of Understanding](https://www.nyk.dev/blog/learn-with-ai) — Guides. Published: 2026-08-30. AI tutoring removes the difficulties that produce retention. Put the model on the question side and add the friction back at four specific points.
- [Security-Review AI-Written Code](https://www.nyk.dev/blog/security-review-ai-written-code) — Guides. Published: 2026-08-21. Generated code is rarely wrong and frequently incomplete in the same seven places. Audit by trust boundary instead of reading the diff for mistakes.
- [Production Coding-Agent Control: 6 Gates Before You Ship](https://www.nyk.dev/blog/production-coding-agent-control-system) — Guides. Published: 2026-08-20. A six-gate repository model for coding-agent authority, capability admission, isolation, evidence, release approval, and recovery.
- [EU AI Act Traceability for Coding Agents: Mapping Obligations to NIST AI RMF and ISO 42001](https://www.nyk.dev/blog/eu-ai-act-traceability-coding-agents) — Guides. Published: 2026-08-11. The Omnibus deferred high-risk obligations to December 2027 and left the record-keeping duty untouched. A deferral is not relief, because the evidence has to exist before the obligation lands.
- [How to Actually Ship OSS Like a Product Without a Second Job](https://www.nyk.dev/blog/ship-oss-like-a-product) — Guides. Published: 2026-08-09. 14.5k+ stars and cold visitors still bounce on two install paths and zero verify. Treat the repository as a product surface: one install path, one proof, one next step.
- [How to Actually Debug Agent Failures: 5 Harness Checks a Bigger Model Cannot Fix](https://www.nyk.dev/blog/agent-harness-failure-matrix) — Guides. Published: 2026-08-09. Lost state, dirty environments, unsafe tools, missing graders, duplicate work. Five failures that survive a model upgrade, and the harness checks that catch them.
- [Agent Governance and Traceability for a Regulated Engineering Org](https://www.nyk.dev/blog/agent-governance-traceability-regulated-org) — Agent systems. Published: 2026-08-09. Which agent changed this, under whose authority, and what did it see? Most teams cannot answer, and the EU AI Act's six-month log retention is the least of why that matters.
- [Claude vs Codex vs Cursor vs Grok: the $200 Wrong-Seat Playbook](https://www.nyk.dev/blog/claude-vs-codex-vs-cursor-vs-grok) — Guides. Published: 2026-08-09. Four coding products keep getting ranked as one model contest. It is a routing failure: an interface, two harnesses, and a model path. Pick the seat, then pick the brand.
- [Why Agent Memory Is Not a Vector Database: The Decision Ledger Playbook in 8 Rules](https://www.nyk.dev/blog/agent-memory-is-not-a-vector-database-decision-ledger) — Agent systems. Published: 2026-08-09. Teams ask for memory and buy an embedding store. An index cannot say what the agent decided last time or whether it was allowed to. A ledger can, and the approval gate comes first.
- [How to Version the Claude System Prompt That Governs Production Before It Breaks](https://www.nyk.dev/blog/system-prompt-is-untested-production-code) — Guides. Published: 2026-08-09. Every other harness component is engineered. The prompt is curated — no diff, no owner, no test, no record of why the last sentence was added. Six checks that close the gap.
- [Why 10 Passing Tests Still Fail My Coding-Agent Merge Gate: 6 Release Checks Before Shipping](https://www.nyk.dev/blog/coding-agent-release-receipt) — Agent systems. Published: 2026-08-07. Ten tests passed in 3.227 seconds and the change still was not merge-ready. The six questions a handoff must answer before an agent-authored patch ships.
- [Your First AI Agent Should Do One Boring Job (7-Step Builder's Roadmap)](https://www.nyk.dev/blog/when-not-to-use-an-agent) — Agent systems. Published: 2026-08-05. The first useful agent I built was not a digital employee.
- [The Production AI Stack for Solo Builders in 2026](https://www.nyk.dev/blog/production-ai-stack-solo-builders-2026) — Agent systems. Published: 2026-08-04. An agent can pass its tests and still ship a silent regression. The model completed its task. The production system failed.
- [Reason Slow. Execute Fast.](https://www.nyk.dev/blog/reason-slow-execute-fast) — Engineering. Published: 2026-08-03. One system asks an LLM to read filings. Another waits for an LLM before it submits an order. Only one puts the model where uncertainty is useful.
- [Production Checklist for a Vibe-Coded Web App](https://www.nyk.dev/blog/production-checklist-vibe-coded-web-app) — Guides. Published: 2026-07-22. A release receipt for clean builds, bad inputs, production journeys, security boundaries, mobile QA, rollback, and post-launch monitoring.
- [How to Debug an AI-Written Next.js App](https://www.nyk.dev/blog/debug-ai-written-nextjs-app) — Guides. Published: 2026-07-22. Classify client, server, build, data, and environment failures; trace one request; isolate the responsible boundary; and prove the repair.
- [AGENTS.md for Next.js Projects: A Production Template](https://www.nyk.dev/blog/agents-md-nextjs-web-projects) — Guides. Published: 2026-07-22. A scoped repository instruction template for Server Components, client boundaries, mutations, verification commands, browser QA, and security.
- [Technical SEO for JavaScript Websites](https://www.nyk.dev/blog/technical-seo-javascript-websites) — Guides. Published: 2026-07-22. Inspect raw HTML, status codes, canonicals, robots, sitemaps, rendered content, structured data, internal links, and Search Console signals.
- [Small-Business Agent Setup: Intake, Follow-Up, and Stop Rules](https://www.nyk.dev/blog/small-business-agent-setup) — Guides. Published: 2026-07-22. Design one bounded agent workflow with structured intake, durable states, narrow tools, payload-bound approvals, retry limits, and receipts.
- [Company Website Launch Checklist: Conversion, SEO, AI Discovery, and Speed](https://www.nyk.dev/blog/company-website-launch-checklist) — Guides. Published: 2026-07-22. A production launch gate for positioning, website UX, identity, contact delivery, technical SEO, AI-readable facts, performance, analytics, and handoff.
- [Solana RPC Failover Without Duplicate Transactions](https://www.nyk.dev/blog/solana-rpc-failover-without-duplicate-transactions) — Guides. Published: 2026-07-22. Rebroadcast one signed transaction across providers, track ambiguous timeouts, and replace it only after blockhash expiration is proven.
- [Build a Safe Solana Transaction Tool for AI Agents](https://www.nyk.dev/blog/safe-solana-transaction-tool-ai-agents) — Guides. Published: 2026-07-22. Put a deterministic intent policy between an AI agent and your Solana signer, with simulation receipts, approval binding, spend caps, and idempotent retries.
- [Stop stacking apps. Build a personal control plane for life, ventures, and work](https://www.nyk.dev/blog/personal-control-plane) — Agent systems. Published: 2026-07-19. Most ambitious people do not fail from laziness.
- [Monitor Solana Transactions Without Missing Events](https://www.nyk.dev/blog/solana-transaction-monitoring-no-missed-events) — Guides. Published: 2026-07-16. Build a live-plus-backfill Solana monitor with durable cursors, overlap recovery, idempotent writes, and explicit commitment handling.
- [Agent Rules vs Skills vs Hooks vs Subagents](https://www.nyk.dev/blog/agent-rules-skills-hooks-subagents-guide) — Guides. Published: 2026-07-16. Choose the cheapest reliable control for persistent context, repeated workflows, lifecycle checks, and isolated agent work.
- [AGENTS.md Guide: What to Include and What to Leave Out](https://www.nyk.dev/blog/agents-md-guide) — Guides. Published: 2026-07-16. Write repository instructions that give coding agents verified commands, safe boundaries, local context, and a clear definition of done.
- [Build a Multi-Agent Review Loop, Not an Agent Swarm](https://www.nyk.dev/blog/multi-agent-review-loop) — Agent systems. Published: 2026-07-16. Use independent positions, shared evidence, one bounded disagreement, and a named authority for decisions that deserve a second reasoning path.
- [Why Your Claude Agent Repeats the Same Mistakes](https://www.nyk.dev/blog/journal-pattern-claude-agent) — Agent systems. Published: 2026-07-16. Build a tested agent journal that turns verified observations into scoped rules and retrieves them before the next matching decision.
- [Hermes Agent Session Management: Resume, Compress, Branch, and Prune](https://www.nyk.dev/blog/hermes-agent-session-management-guide) — Guides. Published: 2026-07-15. A working session protocol for Hermes Agent: name threads, resume the right lineage, compress context, branch experiments, export receipts, and prune safely.
- [Hermes Agent Beginner Guide: Build a Local Agent You Can Operate](https://www.nyk.dev/blog/hermes-agent-beginner-guide) — Guides. Published: 2026-07-15. Install Hermes Agent, choose a provider, set project context, add one skill, and finish with a session you can resume and verify.
- [OpenClaw Beginner Guide: Gateway, Workspace, Memory, and Skills](https://www.nyk.dev/blog/openclaw-beginner-guide) — Guides. Published: 2026-07-15. A cautious OpenClaw setup for vibe coders: run the Gateway, define a workspace, separate sessions, curate memory, and inspect skills before installation.
- [Claude Cowork for Vibe Coding: Files, Projects, Permissions, and Handoff](https://www.nyk.dev/blog/claude-cowork-vibe-coding-guide) — Guides. Published: 2026-07-15. Use Claude Cowork for research, documents, local files, and long-running knowledge work while keeping code verification in a real repository loop.
- [Hermes Agent vs OpenClaw vs Claude Cowork: Pick the Right Agent Host](https://www.nyk.dev/blog/hermes-agent-vs-openclaw-vs-claude-cowork) — Guides. Published: 2026-07-15. A decision matrix for local agent operators: choose Hermes for terminal-first sessions, OpenClaw for gateway-based assistants, or Cowork for managed knowledge work.
- [Single Agent + Skills vs Multi-Agent Teams: When the Committee Loses](https://www.nyk.dev/blog/single-agent-skills-beat-multi-agent-teams) — Guides. Published: 2026-07-14. Most people add more agents when work gets hard. Compile roles into a skill library, keep one owner of the merge, and only go multi-agent for parallel research.
- [Install Agent Skills Without Getting Owned: A Trust Gate Checklist](https://www.nyk.dev/blog/install-agent-skills-without-getting-owned) — Guides. Published: 2026-07-14. Skills are runbooks with teeth. Pre-install and pre-publish checklists so you do not treat stars as security reviews.
- [Memory, Skills, Rules: The Compression Spectrum for Agent Operators](https://www.nyk.dev/blog/memory-skills-rules-compression-spectrum) — Guides. Published: 2026-07-14. Stop pasting chat dumps into every session. Promote experience into journals, skills, and hard rules with a weekly prune.
- [The First Hour After You Post on X: Staff the Thread or Lose It](https://www.nyk.dev/blog/first-hour-after-you-post-on-x) — Guides. Published: 2026-07-14. Publish is half the job. A first-hour staff card for replies, link hygiene, and conversation quality - without inventing algorithm weights.
- [Solana Agent Ops: Devnet Checklist Before You Hand Over Keys](https://www.nyk.dev/blog/solana-agent-ops-devnet-checklist) — Guides. Published: 2026-07-14. Simulate-first, least privilege, human gates on sends. A skill-shaped ops card for agents that touch Solana without demo-day regret.
- [Solana Streaming Paths: Poll → Webhook → WSS → gRPC → Shred](https://www.nyk.dev/blog/solana-streaming-paths-decision-matrix) — Guides. Published: 2026-07-14. Most teams overbuy streams or underbuy completeness. A six-row ladder and decision matrix for app, indexer, and trading workloads.
- [Your AI Agent Didn't Get Dumber. Your Context Did.](https://www.nyk.dev/blog/your-context-did-not-your-agent) — Agent systems. Published: 2026-07-13. You know the feeling. The agent that nailed everything this morning is now making mistakes a beginner would not make. It just got noticeably worse over the sess…
- [How to Actually Package Agent Workflows as Hermes Plugins - Full Course](https://www.nyk.dev/blog/hermes-plugins-app-store-moment) — Agent systems. Published: 2026-07-13. Most people still judge agents by model quality. Packaging workflows as Hermes plugins is the hard part: contracts, distribution, and operator loops.
- [Fable Masterclass: stop spending Mythos-class intelligence like a chat app](https://www.nyk.dev/blog/fable-masterclass-mythos-spend) — Agent systems. Published: 2026-07-13. They will route every task to the most expensive brain in the room.
- [How to Actually Run a Claude Memory Stack: 90 Days of Receipts](https://www.nyk.dev/blog/claude-obsidian-memory-stack-90d) — Agent systems. Published: 2026-07-13. Most people treat Claude memory like a bigger notebook.
- [The 4 Layers of AI Work: Prompt Engineering Is Only Layer 1](https://www.nyk.dev/blog/4-layers-of-ai-work) — Agent systems. Published: 2026-07-13. The argument over prompt engineering versus context engineering is already stale.
- [How to Become an AI-Native Developer (Without a CS Degree Cosplay)](https://www.nyk.dev/blog/how-to-become-ai-native-developer) — Guides. Published: 2026-07-13. AI-native is not ChatGPT tabs. It is specs, tools, verification, and memory - a one-week starter OS for developers.
- [Claude Code Skills for Beginners: Behavior Contracts That Do Not Rot](https://www.nyk.dev/blog/claude-code-skills-beginner-guide) — Guides. Published: 2026-07-13. Skills are not longer prompts. A beginner template, auto-invoke mental model, and monthly prune so your skill library stays green.
- [Agent OS for Solo Developers: Skills, Memory, Jobs, Handoff](https://www.nyk.dev/blog/agent-os-for-solo-developers) — Guides. Published: 2026-07-13. You do not need a platform team. A weekend architecture for solo builders who want agents that compound without chaos.
- [Solana + AI Agents: A Beginner Stack That Survives Devnet](https://www.nyk.dev/blog/solana-ai-agents-beginner-stack) — Guides. Published: 2026-07-13. Harness discipline meets chain hygiene: simulate-first tools, gated writes, context packs, and a two-week learning path.
- [Vibe Coding Stack Mid-2026: Models, Providers, and When to Switch](https://www.nyk.dev/blog/vibe-coding-stack-mid-2026) — Guides. Published: 2026-07-13. As-of July 2026 field map: Claude Code, Codex, Cursor, Gemini CLI - what each is for, what rot looks like, and a default stack that survives real repos.
- [The Vibe Coding Verification Loop: Ship Fast Without Quietly Breaking Everything](https://www.nyk.dev/blog/vibe-coding-verification-loop) — Guides. Published: 2026-07-13. Output is free now. Proof is not. A session checklist that turns agent speed into diffs you can merge.
- [Multi-Model Routing for Vibe Coders: Cheap Loops, Expensive Truth](https://www.nyk.dev/blog/vibe-coding-multi-model-routing) — Guides. Published: 2026-07-13. Stop sending every prompt to the frontier. Route plan / implement / review / audit by cost and failure mode.
- [Mission Control v2.1.0 Is Out - Self-Hosted Mission Control for AI Agents](https://www.nyk.dev/blog/mission-control-v2-1-0-is-out-self-hosted-mission-) — Agent systems. Published: 2026-07-07. Self-hosted mission control for AI agent fleets: what v2.1.0 is for, how to install the mental model, and the operator checklist after clone.
- [4 Layers Every Production AI Agent Needs](https://www.nyk.dev/blog/4-layers-every-production-ai-agent-needs-text-only) — Agent systems. Published: 2026-07-07. Model, tools, context, harness - the compact production checklist before you trust an agent with real work.
- [Context Limits Are the Silent Killer of AI Agents](https://www.nyk.dev/blog/context-limits-are-the-silent-killer-of-ai-agents) — Agent systems. Published: 2026-07-07. Why agents go sloppy mid-session without a clean error - and the four practices that stop silent context death.
- [How I Run Claude Agents on 1M Context Without Drift](https://www.nyk.dev/blog/claude-agents-1m-context-without-drift) — Engineering. Published: 2026-03-15. The problem was never the window size. It was context rot - performance degrades as conversations fill, not because the model forgets, but because signal drowns in noise.
- [Vibe Code a DeFi Quant Bot Without Blowing Up](https://www.nyk.dev/blog/vibe-code-defi-quant-bot) — Guides. Published: 2026-03-14. Most people think the hard part is writing the bot. It isn't. The hard part is building risk rules that the bot cannot negotiate with.
- [If You Have Multiple Interests, Stop Trying to Pick One](https://www.nyk.dev/blog/multiple-interests-stop-picking-one) — Strategy. Published: 2026-03-13. Pick one identity and repeat it for 40 years - that model breaks when the world rewrites itself every quarter. Your curiosity is raw material, not a weakness.
- [The Ultimate Vibe Coding Beginner's Guide](https://www.nyk.dev/blog/ultimate-vibe-coding-beginners-guide) — Guides. Published: 2026-03-12. AI raises output and quietly raises rework: GitClear measured refactoring collapsing from 24.1% to 9.5% of changed lines. The failure pattern, why it happens, and the checklist to survive it.
- [Context Engineering Is the Only Engineering That Matters Now](https://www.nyk.dev/blog/context-engineering-only-engineering) — Engineering. Published: 2026-03-11. An arXiv paper documented a Claude Code project with 26,000 lines of context architecture - more instructions than actual code. Their agents stopped hallucinating.
- [The Harness Is the Product. The Model Never Was](https://www.nyk.dev/blog/the-harness-is-the-product) — Agent systems. Published: 2026-03-10. LangChain jumped from Top 30 to Top 5 by changing zero model parameters. The teams shipping production agents have the best harness, not the best model access.
- [Claude + Obsidian: The Memory Stack That Compounds](https://www.nyk.dev/blog/claude-obsidian-memory-stack) — Engineering. Published: 2026-03-09. Teams burn 30-40 minutes per Claude session re-explaining what they already knew. A three-tier memory system that makes output compound instead of plateau.
- [Your Next Buyer Won't Fill Out a Form. It Will Hit an MCP Endpoint With a Budget.](https://www.nyk.dev/blog/next-buyer-mcp-endpoint) — Growth. Published: 2026-03-07. Visa reported a 4,700% jump in AI-driven retail traffic. Within 12 months, MCP buying shifts revenue upstream. Most GTM teams still optimize for humans.
- [Why AI Needs Crypto (And How to Sell to Agents in a Shifting Market)](https://www.nyk.dev/blog/why-ai-needs-crypto) — Strategy. Published: 2026-02-27. Most people frame AI vs crypto as a rivalry. Wrong framing. If AI is the intelligence layer, crypto is becoming the transaction layer for autonomous software.
- [The Claude + Codex Local Stack Playbook That Cut One Loop by 78.5%](https://www.nyk.dev/blog/claude-codex-local-stack-playbook) — Engineering. Published: 2026-02-20. Updated: 2026-08-05. Everyone says 'fix prompts' - in practice, the bottleneck is shell init, hook chains, and context drift. A measured pass cut zsh startup from 1.794s to 0.386s.

## Canonical indexes

- [What is an Agent OS](https://www.nyk.dev/agent-os): category definition - the production control layer for coding agents - hubbing the agent-systems guides and every product.
- [Portfolio and venture stages](https://www.nyk.dev/ventures)
- [Consulting packages](https://www.nyk.dev/consulting)
- [Self-serve products](https://www.nyk.dev/products)
- [Accessibility](https://www.nyk.dev/accessibility)
- [Selected work and role labels](https://www.nyk.dev/work)
- [Open-source catalog](https://www.nyk.dev/oss)
- [Writing and guides](https://www.nyk.dev/blog)
- [Short machine-readable index](https://www.nyk.dev/llms.txt)
- [XML sitemap](https://www.nyk.dev/sitemap.xml)