{
  "schema_version": "1.0",
  "publisher": "AI Release",
  "article": {
    "title": "Memory for Long-Running AI Agents: Runtime, Not Prompt",
    "canonical_url": "https://ai-release.org/posts/pamyat-dolgozhivuschih-ii-agentov-rantaym-a-ne-prompt-e1acdf26_en.html",
    "language": "en",
    "article_type": "analysis",
    "published_at": "2026-11-18T00:00:00+00:00",
    "updated_at": "2026-11-18T00:00:00+00:00"
  },
  "summary": "In a Hugging Face Forum thread (20 replies, 2,555 views, 15 likes), topic starter Suhebmultani asked how to design memory for an agent that runs for a long time — a chatbot or an assistant. The conversation quickly boiled down to a single question: who in the system is responsible for memory — the model or the code around it?",
  "topics": [
    "AI",
    "Agents",
    "memory"
  ],
  "entities": [
    {
      "id": "E-SRC",
      "name": "discuss.huggingface.co",
      "type": "source",
      "description": "source domain",
      "url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584"
    },
    {
      "id": "E-T1",
      "name": "AI",
      "type": "topic",
      "description": "publication tag",
      "url": null
    },
    {
      "id": "E-T2",
      "name": "Agents",
      "type": "topic",
      "description": "publication tag",
      "url": null
    },
    {
      "id": "E-T3",
      "name": "memory",
      "type": "topic",
      "description": "publication tag",
      "url": null
    },
    {
      "id": "E-AIR",
      "name": "AI Release",
      "type": "publisher",
      "description": "published the article",
      "url": "https://ai-release.org/"
    }
  ],
  "claims": [
    {
      "id": "AR-PAMYATDOLG-001",
      "statement": "In a Hugging Face Forum thread (20 replies, 2,555 views, 15 likes), topic starter Suhebmultani asked how to design memory for an agent that runs for a long time — a chatbot or an assistant.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-001",
      "evidence_excerpt": "In a Hugging Face Forum thread (20 replies, 2,555 views, 15 likes), topic starter Suhebmultani asked how to design memory for an agent that runs for a long time — a chatbot or an assistant."
    },
    {
      "id": "AR-PAMYATDOLG-002",
      "statement": "The conversation quickly boiled down to a single question: who in the system is responsible for memory — the model or the code around it?",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-002",
      "evidence_excerpt": "The conversation quickly boiled down to a single question: who in the system is responsible for memory — the model or the code around it?"
    },
    {
      "id": "AR-PAMYATDOLG-003",
      "statement": "Practitioners answer \"the code,\" but as soon as things get concrete — the database, serialization, forgetting criteria — the thread stalls at architectural diagrams.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-003",
      "evidence_excerpt": "Practitioners answer \"the code,\" but as soon as things get concrete — the database, serialization, forgetting criteria — the thread stalls at architectural diagrams."
    },
    {
      "id": "AR-PAMYATDOLG-004",
      "statement": "Pimpcat-AU sets the tone with a blunt formula: \"Do not make the prompt the memory system.\" He suggests keeping the full history, files, tool outputs, user preferences, decisions made, and task state in a database outside the model, feeding only the relevant facts for the current step into the context.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-004",
      "evidence_excerpt": "Pimpcat-AU sets the tone with a blunt formula: \"Do not make the prompt the memory system.\" He suggests keeping the full history, files, tool outputs, user preferences, decisions made, and task state in a database outside the model, feeding only the relevant facts for the current step into the context."
    },
    {
      "id": "AR-PAMYATDOLG-005",
      "statement": "In a second message he puts it even more bluntly: \"The model is not the memory system.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-005",
      "evidence_excerpt": "In a second message he puts it even more bluntly: \"The model is not the memory system."
    },
    {
      "id": "AR-PAMYATDOLG-006",
      "statement": "The runtime is.\" In his view, the decision about what becomes long-term memory is always made by the application; an LLM can only suggest candidates for memorization — it shouldn't write persistent state without rules around it.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-006",
      "evidence_excerpt": "The runtime is.\" In his view, the decision about what becomes long-term memory is always made by the application; an LLM can only suggest candidates for memorization — it shouldn't write persistent state without rules around it."
    },
    {
      "id": "AR-PAMYATDOLG-007",
      "statement": "marksml (Michael) is the thread's main skeptic.",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-007",
      "evidence_excerpt": "marksml (Michael) is the thread's main skeptic."
    },
    {
      "id": "AR-PAMYATDOLG-008",
      "statement": "He agrees with the architecture but steers the conversation toward engineering: \"Which part of the agent decides what to keep and what to through away aka forget/ignore?\" Who decides — the application logic or the LLM \"brain\"?",
      "claim_type": "reported_fact",
      "confidence": "not_assessed",
      "source_type": "publication source",
      "verification_status": "reported_unverified",
      "source_url": "https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "article_anchor": "#ai-claim-008",
      "evidence_excerpt": "He agrees with the architecture but steers the conversation toward engineering: \"Which part of the agent decides what to keep and what to through away aka forget/ignore?\" Who decides — the application logic or the LLM \"brain\"?"
    }
  ],
  "relationships": [
    {
      "subject": "discuss.huggingface.co",
      "predicate": "reports on",
      "object": "Memory for Long-Running AI Agents: Runtime, Not Prompt",
      "claim_id": "AR-PAMYATDOLG-001"
    },
    {
      "subject": "AI Release",
      "predicate": "publishes",
      "object": "Memory for Long-Running AI Agents: Runtime, Not Prompt",
      "claim_id": "AR-PAMYATDOLG-001"
    }
  ],
  "questions_and_answers": [
    {
      "question": "What is this article about?",
      "answer": "Memory for Long-Running AI Agents: Runtime, Not Prompt. In a Hugging Face Forum thread (20 replies, 2,555 views, 15 likes), topic starter Suhebmultani asked how to design memory for an agent that runs for a long time — a chatbot or an assistant.",
      "claim_ids": [
        "AR-PAMYATDOLG-001"
      ],
      "article_anchor": "#ai-knowledge"
    },
    {
      "question": "What happened?",
      "answer": "In a Hugging Face Forum thread (20 replies, 2,555 views, 15 likes), topic starter Suhebmultani asked how to design memory for an agent that runs for a long time — a chatbot or an assistant. The conversation quickly boiled down to a single question: who in the system is responsible for memory — the model or the code around it? Practitioners answer \"the code,\" but as soon as things get concrete — the database, serialization, forgetting criteria — the thread stalls at architectural diagrams.",
      "claim_ids": [
        "AR-PAMYATDOLG-001",
        "AR-PAMYATDOLG-002",
        "AR-PAMYATDOLG-003"
      ],
      "article_anchor": "#ai-knowledge"
    },
    {
      "question": "Who is the source?",
      "answer": "discuss.huggingface.co: https://discuss.huggingface.co/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584",
      "claim_ids": [],
      "article_anchor": "#ai-knowledge"
    },
    {
      "question": "What remains unknown?",
      "answer": "The article reports the cited source; independent verification and full details are not provided.",
      "claim_ids": [],
      "article_anchor": "#ai-knowledge"
    }
  ],
  "limitations": [
    "No independent verification is provided.",
    "Full details, dates and official documents may be missing.",
    "The article reports a source rather than being the primary source."
  ],
  "citation": {
    "preferred_citation": "AI Release. «Memory for Long-Running AI Agents: Runtime, Not Prompt». 2026-11-18. https://ai-release.org/posts/pamyat-dolgozhivuschih-ii-agentov-rantaym-a-ne-prompt-e1acdf26_en.html",
    "canonical_url": "https://ai-release.org/posts/pamyat-dolgozhivuschih-ii-agentov-rantaym-a-ne-prompt-e1acdf26_en.html",
    "publisher": "AI Release"
  }
}