{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreidmbpjcg3uiplrs67hybvrwpykxnqilsaxx4p7cstzruww4uacavi",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mlqvummvraz2"
  },
  "path": "/t/how-do-you-design-memory-systems-for-long-running-ai-agents/175584#post_10",
  "publishedAt": "2026-05-13T17:09:35.000Z",
  "site": "https://discuss.huggingface.co",
  "textContent": "the rudimentry system i started useing was a multi-artifact system\n\nLayer 1\nwhen i put in a prompt, that gets saved as a physical artifact. when the system responds, that is also a physical artifact, like a log file.\nthe prompt is timestamped and labled for the conversation, so that the 2 are cross referenceable.\n\nLayer 2\ni use artifacts like Project roadmaps and ‘Notes’ files.\nthe projects - nested roadmaps alow for outlining projects, that can be expandable. the notes are what the AI uses to track its progress and thought process. and these 2 are saved with ttimestamps and unique identifiers teing them back to the conversation that they are associated with.\n\nthere are other ways this system can be expanded.\n\nbut the reasonings behind this system are simple.\n\n  1. text files are microscopic. the only real danger is file and name organization.\n  2. a comprehensive record is kept, but the system can nominally reference the notes file for shorthand, but if it needs more context it can go all the way back to the actual conversation.\n\n\n\nbut i get the impression there are alot of different ways this type of memory issue can be adressed.",
  "title": "How do you design memory systems for long-running AI agents?"
}