Multi-Agent Collaboration for Team and Friends
TypeScript agent-automationagent-orchestrationagent-workflowai-agentsai-collaborationai-memoryai-workforceautonomous-agentsclaude-codeclaude-code-skillscodexcoding-agentmulti-agentmultiple-agentone-person-businessone-person-companyone-person-teamopencodesolopreneur
seo(blog): revise personal-ai-company guide and diagrams (#398)
LindsayLiu7779ed9fc679263
diff --git a/src/web/public/blog/personal-ai-company/comparison.webp b/src/web/public/blog/personal-ai-company/comparison.webpindex eace90b62..c9cfaf348 100644Binary files a/src/web/public/blog/personal-ai-company/comparison.webp and b/src/web/public/blog/personal-ai-company/comparison.webp differdiff --git a/src/web/public/blog/personal-ai-company/hero.webp b/src/web/public/blog/personal-ai-company/hero.webpindex 388a18314..fd5425bbb 100644Binary files a/src/web/public/blog/personal-ai-company/hero.webp and b/src/web/public/blog/personal-ai-company/hero.webp differdiff --git a/src/web/public/blog/personal-ai-company/timeline.webp b/src/web/public/blog/personal-ai-company/timeline.webpindex 87dbc25be..a6fbcb316 100644Binary files a/src/web/public/blog/personal-ai-company/timeline.webp and b/src/web/public/blog/personal-ai-company/timeline.webp differdiff --git a/src/web/src/content/personal-ai-company.mdx b/src/web/src/content/personal-ai-company.mdxindex bc78eb0b8..79937dc67 100644--- a/src/web/src/content/personal-ai-company.mdx+++ b/src/web/src/content/personal-ai-company.mdx@@ -1,11 +1,12 @@export const metadata = {slug: "personal-ai-company",- title: "How to Build a One-Person Business With AI Agents",+ title: "How to Build a One-Person Company With AI Agents",date: "2026-06-10",+ dateModified: "2026-07-23",author: "Alook Team",excerpt:- "Build a one-person business with AI agents through repeatable workflows, defined handoffs, limited access, and human approval for high-impact work.",- readingTime: "6 min read",+ "Build a one-person company with AI agents through repeatable workflows, defined handoffs, limited access, and human approval for high-impact work.",+ readingTime: "7 min read",image: "/blog/personal-ai-company/hero.webp",};@@ -19,7 +20,15 @@ export const jsonLd = [name: "What is a personal AI company?",acceptedAnswer: {"@type": "Answer",- text: "A personal AI company is a one-person business where the founder uses AI agents for defined workflows while keeping responsibility for goals, approvals, and important decisions.",+ text: "A personal AI company is a one-person company where the founder uses AI agents for defined workflows while keeping responsibility for goals, approvals, and important decisions.",+ },+ },+ {+ "@type": "Question",+ name: "How is an AI agent team different from a one-person company?",+ acceptedAnswer: {+ "@type": "Answer",+ text: "A one-person company is the business shape: one founder stays accountable. An AI agent team is the operating pattern inside it, with scoped roles, handoff artifacts, and approval gates. A solo founder can run a small agent team without hiring a traditional department roster.",},},{@@ -27,7 +36,7 @@ export const jsonLd = [name: "Which AI agents does Alook support?",acceptedAnswer: {"@type": "Answer",- text: "Alook currently supports Claude Code, Codex, and OpenCode. It gives these local coding agents roles, email inboxes, task tracking, schedules, and a shared workspace for collaboration.",+ text: "See the current supported agents list on the AI agent team guide. Alook gives local coding agents roles, email inboxes, task tracking, schedules, and a shared workspace for collaboration.",},},{@@ -35,7 +44,7 @@ export const jsonLd = [name: "Can AI agents run a one-person business without human oversight?",acceptedAnswer: {"@type": "Answer",- text: "No. Agents can handle scoped, repeatable work, but the founder should approve customer commitments, financial actions, legal decisions, pricing changes, and other high-impact work.",+ text: "No. Whether you call it a one-person company or a one-person business, agents can handle scoped, repeatable work, but the founder should approve customer commitments, financial actions, legal decisions, pricing changes, and other high-impact work.",},},{@@ -50,29 +59,30 @@ export const jsonLd = [},];-A one-person business does not need an AI version of every department. It needs a small number of reliable workflows that keep moving when the founder is focused elsewhere.+A one-person company does not need an AI version of every department. It needs a few reliable workflows that keep moving when you are heads-down elsewhere.-That is the practical idea behind a personal AI company. You remain responsible for the business. AI agents take ownership of defined work, use the tools you allow, and return results for review. As more agents are added, an orchestration layer keeps their roles, tasks, schedules, and communication connected.+That is the practical idea behind a personal AI company — what many founders also call a one-person business. You stay responsible for the company. Agents take ownership of defined work, use only the tools you allow, and return results for review. When more than one agent is involved, an orchestration layer keeps roles, tasks, schedules, and communication connected.-+-*This diagram is a conceptual example of role specialization. The founder remains responsible for company direction and approval.*+*Conceptual role map: the founder stays human. Agents own scoped workflows and stop at approval boundaries.*-## What is a one-person company AI setup?+## What is a personal AI company?-A one-person company AI setup combines three parts:+Bingran You’s [2026 working definition](https://bingran.ai/one-person-company) puts it plainly: a one-person company is not “a founder with a chatbot.” It is a founder running several agentic surfaces, each with a specific job, glued together by orchestration and rails. In practice the setup has a few moving parts:- **A founder who sets direction.** You choose priorities, define boundaries, and approve decisions with financial, legal, or reputational impact.- **Agents with narrow roles.** Each agent gets a clear responsibility, the context needed to do the work, and access only to relevant tools.- **An orchestration layer.** This coordinates assignments, handoffs, recurring work, and records so the operation does not depend on you forwarding every message.+- **A review habit.** Someone human still checks the outputs that can hurt customers or production.-The goal is not to pretend software is an employee. The goal is to make repeatable work easier to delegate and inspect.+The point is narrower than “AI employees.” Make repeatable work easier to delegate and inspect. Useful starting workflows include a morning brief, turning a bug report into a draft pull request, organizing release notes, researching a lead, or drafting a follow-up for approval — tasks with visible inputs and outputs you can check before they reach a customer or change production.-Useful starting workflows include preparing a morning brief, turning a bug report into a draft pull request, organizing release notes, researching a lead, or drafting a follow-up for approval. These tasks have visible inputs and outputs. You can check the result before it reaches a customer or changes production.+## Start from the workflow, not the job title-## Which AI agents should a solopreneur start with?+A “marketing agent” is too broad to evaluate.-Start with the workflow, not the job title. A “marketing agent” is too broad to evaluate. “Draft a weekly product update from merged pull requests” gives the agent a source, an output, and a review point.+“Draft a weekly product update from merged pull requests” gives the agent a source, an output, and a review point. That is the kind of brief you can actually judge.A good first workflow usually has these properties:@@ -82,52 +92,35 @@ A good first workflow usually has these properties:- A mistake can be caught before it causes harm.- Success can be checked without relying on vague judgments.-Avoid starting with a workflow that requires constant exceptions, sensitive customer decisions, or unrestricted access to business systems. If every run needs a different set of instructions, the process is not ready for delegation.--## Best AI tools for solopreneurs: start with orchestration--There is no universal “2026 stack” for every solopreneur. A consultant, an open-source maintainer, and an online retailer use different systems. Recommending the same payment, analytics, and website tools to all of them would add noise without helping them build a reliable workflow.--The shared requirement is coordination. [AI orchestration](/blog/ai-agent-orchestration) connects roles, assignments, context, and handoffs across agents. [Alook](https://alook.ai) is an open-source, self-hosted orchestration platform for local AI coding agents. It currently supports Claude Code, Codex, and OpenCode.--Alook gives those agents an operating structure:+What should you skip? Constant exceptions, sensitive customer decisions, unrestricted system access, or a process that needs a brand-new brief every run.-- **Roles and an org chart** define which agent owns each kind of work.-- **An `@alook.ai` email address** lets people and agents exchange instructions and replies.-- **Kanban task tracking** records assignments and progress.-- **Calendar scheduling** supports reminders, recurring tasks, and routines.-- **Persistent context and traceability** preserve decisions, instructions, and responses.-- **Local execution** keeps agents and their tool access on your machine while allowing remote access through Alook.+If you are running a one-person business today, start from a scoped workflow with a clear review point, not from a stack of job titles.-You can currently get started with Alook for free. It is also licensed under Apache 2.0 and can be self-hosted. That describes Alook's current availability; it is not a claim about the total cost of running a business or using third-party models and services.+## How to build a one-person company with AI agents-+### 1. Write Friday’s newsletter as a contract-*This progression is illustrative. The useful number of agents depends on the workflows you have tested and connected.*+Name the trigger, inputs, steps, output, and approval point. “Handle content” fails that test. “Prepare a draft newsletter from these approved sources every Friday” is specific enough to run and reject.-## How to build a one-person business with AI agents+### 2. Give one agent a narrow inbox and tool set-### 1. Choose one recurring workflow+Assign a role and only the tools that workflow needs. A research brief may need approved sources and a document destination — not billing or production access.-Write down the trigger, inputs, steps, output, and approval point. “Handle content” is too broad. “Prepare a draft newsletter from these approved sources every Friday” is specific enough to test.+### 3. Say where the draft goes next-### 2. Assign one agent and limited access+Decide whether the result returns to you, creates a task for another agent, or waits in a draft state. Incomplete workflows leave nobody sure what happens after the first output.-Give the agent a role and only the tools needed for that workflow. A research task may need approved web sources and a document destination. It does not need access to billing or production infrastructure.+### 4. Break the Friday run on purpose-### 3. Define the handoff+Use representative inputs first. Then test missing data, tool errors, ambiguous requests, and drafts that should be rejected. Keep the work supervised until failure behavior is predictable.-Decide where the result goes next. It might return to you for approval, create a task for another agent, or wait in a draft state. A workflow is incomplete if nobody knows what should happen after the first output.+### 5. Put it on the calendar only after review-### 4. Test normal cases and failures+Once the unscheduled version produces useful results, add a schedule or trigger. Review history when the underlying business process changes. Add another agent only when the next handoff is clear — team size is a poor measure of progress.-Run the workflow with representative inputs. Then test missing data, tool errors, ambiguous requests, and outputs that should be rejected. Keep the work supervised until the failure behavior is predictable.+-### 5. Schedule it only after review--Once the workflow produces useful results consistently, add a schedule or trigger. Review its history and adjust the instructions when the underlying business process changes.--Add another agent only when the next handoff is clear. More agents create more coordination work, so team size is a poor measure of progress.+*Illustrative maturity path. Grow by tested workflows and handoffs, not by inventing department titles.*## Where human approval belongs@@ -139,38 +132,48 @@ Some decisions should stay with the founder even when an agent prepares the work- Publishing sensitive claims- Deploying high-risk production changes-Agents can gather context, draft options, and carry out approved steps. You remain accountable for the decision. This boundary protects customers and gives you a clean audit trail when something needs review.+On [scalable.dk](https://scalable.dk/posts/agentic-sdlc-in-production/), a solo founder describes agent pipelines that open draft pull requests and stop there: the routines never merge. That pattern matches the boundary here. Agents can gather context, draft options, and carry out approved steps. You remain accountable for the decision.-## What makes a personal AI company work?+## Chat sessions vs a working personal AI company-The useful distinction is between a collection of chat sessions and an operation with defined ownership.+Isolated agents lose context between tasks and force you to relay information manually. Coordinated agents can work from recorded decisions, send work to a specific teammate, and leave a trace of what happened. That helps. Oversight still matters.-Isolated agents lose context between tasks and force you to relay information manually. Coordinated agents can work from recorded decisions, send work to a specific teammate, and leave a trace of what happened. That reduces routing work, but it does not remove the need for clear instructions or oversight.+Vendor tooling inside one product can help peers coordinate. Anthropic’s [Claude Code agent teams](https://code.claude.com/docs/en/agent-teams) document lead/teammate sessions with a shared task list and messaging, and note that the mode is experimental. A personal AI company still sits one layer up: which workflows you own, which artifacts move between stages, and where you approve.-Start with one workflow that has a visible result. Make it dependable. Then connect a second workflow where a real handoff exists. A personal AI company grows from tested processes, not from adding impressive agent titles to an org chart.+-+*Conceptual contrast only: routing burden and ownership, not salaries, headcount, or pricing.*-*This image is a conceptual comparison. The figures shown are hypothetical examples and do not represent Alook pricing, actual salaries, operating costs, agent limits, or guaranteed working hours.*+## Keep coordination off your calendar-## Personal AI company FAQ+[AI orchestration](/blog/ai-agent-orchestration) is the shared requirement across solopreneur stacks. Consultants, open-source maintainers, and retailers use different tools, yet they still need roles, assignments, context, and handoffs that do not collapse into manual copy-paste.++A **one-person company** is the shape of the business: one founder stays accountable. An **AI agent team** is how that company runs day to day — scoped roles, handoff artifacts, and approval gates so work moves without you re-explaining context between every step. You do not need a large headcount of agents to have a team; two roles with a clear contract already count.++For the roles, contracts, and failure modes in detail, read **[How to Build an AI Agent Team](/blog/ai-agent-team)**. Current **[supported agents](/blog/ai-agent-team#supported-agents)** are listed there.-**What is a personal AI company?**+## Building this with Alook++[Alook](https://alook.ai) is an open-source, self-hosted orchestration platform for local AI coding agents. It gives agents an operating structure: roles and an org chart, individual `@alook.ai` email addresses, Kanban tracking, calendar scheduling, persistent context, traceability, and local execution on your machine. You can currently get started with Alook for free under Apache 2.0, including self-hosting.++That describes Alook’s current availability. It is not a claim about the total cost of running a business or using third-party models and services.++## Personal AI company FAQ-It is a one-person business where AI agents handle defined workflows while the founder controls goals, access, and important approvals.+**What is a personal AI company?** It is a one-person company where AI agents handle defined workflows while the founder controls goals, access, and important approvals.-**Which AI agents does Alook support?**+**How is an AI agent team different from a one-person company?** The company is the business shape. The agent team is the operating pattern inside it: roles, handoffs, and approval gates. A solo founder can run a small agent team without pretending they hired a C-suite.-Alook currently supports Claude Code, Codex, and OpenCode. Cursor, Hermes, and OpenClaw are listed as coming soon.+**Which AI agents does Alook support?** See the current [supported agents](/blog/ai-agent-team#supported-agents) list.-**Can AI agents run the company without human oversight?**+**Can AI agents run a one-person business without human oversight?** No. Keep human approval around customer promises, financial actions, legal decisions, pricing changes, and other high-impact work. That gate stays even when the agents look reliable.-No. Keep human approval around customer promises, financial actions, legal decisions, pricing changes, and other high-impact work.+**Is Alook free to use?** You can currently get started with Alook for free under the Apache 2.0 license, including self-hosting.-**Is Alook free to use?**+## Run one Friday loop before you hire titles-You can currently get started with Alook for free. The project is open source, uses the Apache 2.0 license, and supports self-hosting.+Pick a recurring job you already do alone — the Friday newsletter, the morning brief, or the bug-to-draft-PR path.-## Start with one workflow+Write the artifact you expect, the tools the agent may touch, and the moment you must approve before anything reaches a customer. When that loop is boring in a good way, add the next handoff. Until then, one reliable workflow beats a pretend org chart.-To try Alook, run `npx @alook/app onboard` and follow the setup flow for your local agent environment. You can also visit [Alook.ai](https://alook.ai) to learn about agent email addresses, templates, and the current product.+To connect local coding agents under Alook, run `npx @alook/app onboard`. Templates and product details live at [alook.ai](https://alook.ai).