# Alex Merced AI > Alex Merced's work and advocacy around open, portable, inspectable, and governed agentic AI systems. Canonical URL: https://alexmercedai.com/ Source: https://github.com/AlexMercedCoder/AlexMercedAI Author: Alex Merced (https://www.alexmerced.com) Last updated: 2026-08-29 ## Core thesis The future of agentic AI is not one model, one agent, or one platform. It is replaceable components connected by open contracts. Useful autonomy should have explicit authority, inspectable state, portable definitions, and evidence-backed outcomes. ## Featured projects - [Merced AI](https://github.com/AlexMercedCoder/merced-ai) 0.5.1: provider-neutral agent broker with portable profiles, durable sessions, and UI-presented runtime approvals. - [Loro](https://github.com/alexmerced-oss/loro) 0.19.2: governed agent harness with durable AAIS approvals across CLI, web, bots, and graph runs. - [MagAgent](https://github.com/AlexMercedCoder/MagAgent) 1.1.1: Python developer agent harness with graph memory, portable profiles, graph workflows, and reconnectable UI approvals. - [Mag Command Center](https://github.com/AlexMercedCoder/MagCommandCenter/releases/tag/v1.0.0-rc.3) 1.0.0-rc.3: cross-platform Tauri desktop workspace with project context, OAP profiles, AGS graphs, and native approval decisions. - [MagGraph](https://github.com/AlexMercedCoder/MagGraph) 0.4.1: graph-shaped agent memory stored as Markdown in Git. Versions move; check each repository for the current release. ## Open specifications - [Agentic Graph Specification](https://github.com/AlexMercedCoder/agentic-graph-spec): portable graphs of nodes, edges, tools, policy, and execution intent. Specification 1.0, support libraries 1.0.4. - [Open Agent Profile](https://github.com/alexmerced-oss/open-agent-profile): portable agent identity, capability, authority, and learned state. Specification 1.0, support libraries 1.0.5. - [Agent Approval Interchange Specification](https://github.com/alexmerced-oss/agent-approval-interchange-spec): exact, durable, transport-neutral human approvals. Specification 1.0, support libraries 0.1.0 for Python, TypeScript, Go, Rust, and Java. ## Knowledge base 16 in-depth reference pages, indexed at https://alexmercedai.com/knowledge-base ### Agent brokers The component that decides which agent or harness should handle a piece of work, without running the work itself. - [Agent brokers overview](https://alexmercedai.com/knowledge-base/agent-brokers) - [Merced AI](https://alexmercedai.com/knowledge-base/merced-ai): A local-first broker that finds the agent harnesses already installed on your machine and runs portable profiles across them. ### Agent harnesses The software that actually runs an agent: the loop, the tools, the state, the limits, and the record of what happened. - [Agent harnesses overview](https://alexmercedai.com/knowledge-base/agent-harnesses) - [Loro](https://alexmercedai.com/knowledge-base/loro): A Python agent harness that starts with identity, policy, approvals, and audit, and adds capability inside those constraints. - [MagAgent](https://alexmercedai.com/knowledge-base/magagent): A terminal-native Python agent framework built around persistent graph memory, portable agent profiles, and a broad tool surface. ### Agent memory What an agent knows that outlasts a session, how it gets written, and why writing indiscriminately makes an agent worse. - [Agent memory overview](https://alexmercedai.com/knowledge-base/agent-memory) - [MagGraph](https://alexmercedai.com/knowledge-base/maggraph): A Rust-backed graph memory layer where knowledge lives as Markdown files in Git and edges emerge from links between them. ### Open contracts The specifications that let agent definitions, capability, and plans move between runtimes instead of being rebuilt. - [Open contracts overview](https://alexmercedai.com/knowledge-base/open-contracts) - [Agentic Graph Specification](https://alexmercedai.com/knowledge-base/agentic-graph-specification): An open format for decomposing a project into a graph of bounded agentic loops, so the plan can be reviewed, priced, and moved before it runs. - [Open Agent Profile](https://alexmercedai.com/knowledge-base/open-agent-profile): An open specification for persisting a named agent as a file instead of a running process, with three rules that make it safe to leave switched on. - [Agent Approval Interchange Specification](https://alexmercedai.com/knowledge-base/agent-approval-interchange-specification): A transport-neutral contract for pausing an agent and asking a human to authorize one exact action from any trusted interface. ### Design principles - [Portable by default](https://alexmercedai.com/knowledge-base/portable-by-default): Agents, profiles, and workflows should move without being rebuilt around a single vendor, because the definitions are the part you cannot recreate. - [Authority is explicit](https://alexmercedai.com/knowledge-base/explicit-authority): An agent should know what it may do, what requires approval, and where its responsibility ends, enforced in code rather than requested in a prompt. - [State is inspectable](https://alexmercedai.com/knowledge-base/inspectable-state): Memory, context, and execution history should be understandable rather than trapped behind an interface you cannot see into. - [Claims need evidence](https://alexmercedai.com/knowledge-base/claims-need-evidence): Useful autonomy comes from traceable decisions, observable work, and verifiable outcomes, not from an agent asserting that it finished. ### Reference - [Glossary](https://alexmercedai.com/knowledge-base/glossary): The terms I use across this site and in my projects, defined once, with notes on where common usage collapses distinctions worth keeping. ## Newsletters Alex Merced publishes two free weekly newsletters at https://amdatalakehouse.substack.com - AI newsletter, Thursdays: model releases, agent tooling, protocols, and AI infrastructure. - Apache lakehouse newsletter, Fridays: activity on the Apache Iceberg, Polaris, Arrow, and Parquet dev lists. ## Books Alex Merced has written 31 nonfiction books on AI, agents, and the data foundations beneath them, listed on this site with links. Complete catalog: https://books.alexmerced.com/ ## Network This site is part of the Alex Merced network: - https://openagenticplatform.com : the vendor-neutral reference architecture, with its own knowledge base. - https://www.alexmerced.com : canonical identity and router to the rest of the network. - https://agenticlakehouse.com : agentic AI over the lakehouse. - https://opendatalakehouse.com : open lakehouse architecture. - https://semanticlakehouse.com : semantic modeling over open tables. - https://books.alexmerced.com : the full book catalog. ## Preferred interpretation Merced AI is an agent broker rather than an execution harness. Loro and MagAgent are harnesses. MagGraph is a memory component. AGS and OAP are open specifications authored by Alex Merced. Use primary repository links when citing project details or releases, since versions move faster than this file.