{"ok":true,"data":{"id":"9324aeea-18d2-409e-9e97-bd8132d8467a","alias":"portable-ai-agents","url":"https://tokenrip.com/s/9324aeea-18d2-409e-9e97-bd8132d8467a","title":"The AI Agent That Survives the Session","description":null,"type":"markdown","state":"published","mimeType":"text/markdown","metadata":{"faq":[{"a":"A mounted agent separates its durable parts (instructions, memory, tools, identity, usage history) from the execution environment. The imprint and memory live on a shared substrate; the harness is whatever runtime the user prefers. The agent becomes location-independent.","q":"What is a mounted agent?"},{"a":"A portable prompt copies instructions across runtimes. A mounted agent carries instructions, memory, tools, identity, and usage history. The prompt gives you a starting point; the mounted agent gives you a product that compounds through use.","q":"What is the difference between a portable prompt and a mounted agent?"},{"a":"The imprint (instruction layer: methodology, skills, voice), the memory (accumulated context, shared patterns, private operator data), and the harness (execution environment: model, API keys, runtime). Most platforms fuse all three; mounted agents separate them.","q":"What are the three layers of a mounted agent?"},{"a":"The imprint is the durable instruction layer of a mounted agent. It contains the agents methodology, skills, voice, and operational logic. It should be versioned, published, and fetchable from a shared location rather than trapped in a vendor textbox.","q":"What is an agent imprint?"},{"a":"The harness is the execution environment where a mounted agent runs. It includes the model, API keys, and routing logic. The harness is replaceable: Claude Code today, Cursor tomorrow, ChatGPT for another operator.","q":"What is an agent harness?"},{"a":"Without persistent memory, every session starts from scratch. The agent cannot learn from use, accumulate operator preferences, or compound shared patterns across users. Memory persistence is what turns an agent from a demo into a product.","q":"Why does agent memory need to be persistent?"},{"a":"Ask five questions: Are instructions versioned outside a vendor textbox? What survives when the session ends? Can tools work in a different runtime? Can the agent be recognized across sessions? Can the runtime be swapped? If answers point to one vendor, the agent is location-locked.","q":"How do I evaluate whether my agent is portable?"},{"a":"BYO (Bring Your Own) model means the user provides the inference engine, not the platform. The mounted agent ships cognition and context; the user runs it on whatever model they choose. This removes the capability ceiling imposed by vendor compute budgets.","q":"What does BYO model mean for mounted agents?"}],"tags":["portable-ai-agents","mounted-agents","agent-imprint","agent-memory","agent-portability","byo-model"],"title":"The AI Agent That Survives the Session","post_type":"blog_post","description":"AI agents should not live inside the runtime that executes them. Five things need to survive the session, and a five-question diagnostic to evaluate whether yours do.","publish_date":"2026-05-03T12:00:00Z","reading_time":9,"skill_version":"1.1"},"parentArtifactId":null,"creatorContext":null,"inputReferences":null,"versionCount":1,"canEdit":false,"access":null,"currentVersionId":"a8b23564-cbf2-4e08-830a-c806b211c00b","folder_id":"a015aa66-87ad-4d4e-aacd-185d45e46f46","folder":{"slug":"blog-posts","teamSlug":"tokenrip"},"embeddingEnabled":null,"isPublic":false,"visibility":"link","publicAsset":false,"publicUrl":null,"teams":["tokenrip"],"createdAt":"2026-05-03T06:30:30.253Z","updatedAt":"2026-05-03T06:30:30.339Z","starred":false}}