OpenExecutive, a virtual executive team of 8 open-source AI agents built on Claude, reached 712 points and 472 comments on Hacker News on August 27, 2026. The GitHub account SenteLabsAI publishes the project under the Apache 2.0 license, and the repository holds 1,052 stars and 59 forks.
The project packages 8 executive roles — Chief Strategy Officer, Chief Financial Officer, Chief HR/People Officer, General Counsel, Chief Operating Officer, Chief Marketing Officer, Chief Product Officer, and Board Communications Director — behind one orchestrator that answers in a single executive voice. The repository was created on June 11, 2026. The Hacker News submission carried the title “CEO fired developers to make room for AI. Developers create open source AI CEO.”
Who Built OpenExecutive and Why
A development team that lost its jobs built OpenExecutive and open-sourced it to sell support services around it, according to Hacker News submitter GrumpySciGuy. “Team was let go. Decided to do a startup,” GrumpySciGuy wrote on August 27, 2026, describing the commercial plan as “similar to the Redhat model and other open source tech.”
GrumpySciGuy named no employer, and no commenter in the 472-comment thread identified the company or the executive behind the layoffs. One commenter, cactusplant7374, asked directly which company fired the team and received no reply. GrumpySciGuy also stated no affiliation with SenteLabsAI, the account that owns the repository.
Which 8 Open-Source AI Agents Does OpenExecutive Ship?
OpenExecutive ships 8 open-source AI agents, each scoped to one executive domain: the Chief Strategy Officer handles competitive analysis and OKRs, the CFO handles financial modeling and unit economics, the Chief HR/People Officer handles hiring and compensation, the General Counsel handles contracts and compliance, the COO handles process design and vendor management, the CMO handles go-to-market and PR, the CPO handles roadmap and prioritization, and the Board Communications Director handles board decks and investor relations.
An orchestrator running claude-sonnet-4-6 routes each question to the relevant specialists in parallel, then synthesizes one answer. Four deep-reasoning roles — Chief Strategy Officer, CFO, General Counsel, and Board Communications Director — run on claude-opus-4-7 with extended thinking. Hacker News commenter rmah identified the gap in the lineup: “the set of agents don’t include an AI CEO, just the execs underneath the CEO.”
What OpenExecutive Runs On
OpenExecutive runs on Python 3.11 with FastAPI on the backend and Next.js 15 on the frontend, with ChromaDB for document retrieval and SQLite for episodic memory. The system requires an ANTHROPIC_API_KEY unless the operator routes through OpenRouter or a local OpenAI-compatible server such as Ollama, LM Studio, or vLLM. A background claude-haiku-4-5 pass extracts decisions after every response, so the next session opens with a record of prior advice.
The repository ships 29 evaluation scenarios covering all 8 domains, scored by claude-opus-4-7 across 5 dimensions including persona coherence, routing quality, and actionability. The continuous-integration gate requires an average of 3.5 out of 5. Six interfaces are included: web UI, Slack, email, Telegram, Google Chat, and Discord.
What This Means for Businesses Evaluating AI Agents
OpenExecutive is a readable reference implementation of multi-agent delegation, which is the architecture most commercial agent platforms hide behind a dashboard. Teams comparing packaged alternatives can start with our roundup of the 12 best AI agents for business tasks, which covers the vendors that sell this same routing-and-synthesis pattern as a managed product.
Two caveats apply to any evaluation. OpenExecutive carries no published production deployments, and its advice quality is graded only by its own LLM-as-judge evals, not by outcomes at a real company.
For Context
Our earlier reporting covers the gap between agent ambition and agent deployment: 99% of companies plan AI agents and only 10% ship them, while Salesforce data shows enterprise AI agent use tripled in a year. For the underlying patterns these systems automate, see our breakdown of 15 AI agent examples across industries.
Related Coverage
Interoperability between agent systems is consolidating around one standard, covered in our report on Google A2A joining MCP under the Agentic AI Foundation.
Our Take
OpenExecutive is part protest and part working product, and the working part is the more useful half. A laid-off team shipped, in 11 weeks, a multi-agent system that routes strategy, finance, legal, and operations questions through 8 specialists and returns one coherent answer. That capability is no longer the hard part. Deciding which decisions a company will actually delegate to it is.
