🐦 Twitter Post Details

Viewing enriched Twitter post

@omarsar0

// Continual Learning from Experience and Skills // Skills are so good when you combine them properly with MCP & CLIs. I have found that Skills can significantly improve tool usage of my coding agents. The best way to improve them is to regularly document improvements, patterns, and things to avoid. Self-improving skills don't work that well (yet). Check out this related paper on the topic: It introduces XSkill, a dual-stream continual learning framework. Agents distill two types of reusable knowledge from past trajectories: experiences for action-level tool selection, and skills for task-level planning and workflows. Both are grounded in visual observations. During accumulation, agents compare successful and failed rollouts via cross-rollout critique to extract high-quality knowledge. During inference, they retrieve and adapt relevant experiences and skills to the current visual context. Evaluated across five benchmarks with four backbone models, XSkill consistently outperforms baselines. On Gemini-3-Flash, the average success rate jumps from 33.6% to 40.3%. Skills reduce overall tool errors from 29.9% to 16.3%. Agents that accumulate and reuse knowledge from their own trajectories get better over time without parameter updates. I have now seen two papers this week with similar ideas. Paper: https://t.co/YXrHcJ6Zim Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX

Media 1
Media 2

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2032928526022881399/media_0.jpg",
      "filename": "media_0.jpg"
    },
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2032928526022881399/media_1.png",
      "filename": "media_1.png"
    }
  ],
  "processed_at": "2026-03-14T21:20:43.145774",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2032928526022881399",
  "url": "https://x.com/omarsar0/status/2032928526022881399",
  "twitterUrl": "https://twitter.com/omarsar0/status/2032928526022881399",
  "text": "// Continual Learning from Experience and Skills //\n\nSkills are so good when you combine them properly with MCP & CLIs.\n\nI have found that Skills can significantly improve tool usage of my coding agents.\n\nThe best way to improve them is to regularly document improvements, patterns, and things to avoid.\n\nSelf-improving skills don't work that well (yet).\n\nCheck out this related paper on the topic:\n\nIt introduces XSkill, a dual-stream continual learning framework.\n\nAgents distill two types of reusable knowledge from past trajectories: experiences for action-level tool selection, and skills for task-level planning and workflows.\n\nBoth are grounded in visual observations.\n\nDuring accumulation, agents compare successful and failed rollouts via cross-rollout critique to extract high-quality knowledge. During inference, they retrieve and adapt relevant experiences and skills to the current visual context.\n\nEvaluated across five benchmarks with four backbone models, XSkill consistently outperforms baselines. On Gemini-3-Flash, the average success rate jumps from 33.6% to 40.3%. Skills reduce overall tool errors from 29.9% to 16.3%.\n\nAgents that accumulate and reuse knowledge from their own trajectories get better over time without parameter updates.\n\nI have now seen two papers this week with similar ideas.\n\nPaper: https://t.co/YXrHcJ6Zim\n\nLearn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX",
  "source": "Twitter for iPhone",
  "retweetCount": 0,
  "replyCount": 0,
  "likeCount": 7,
  "quoteCount": 1,
  "viewCount": 413,
  "createdAt": "Sat Mar 14 21:15:05 +0000 2026",
  "lang": "en",
  "bookmarkCount": 6,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2032928526022881399",
  "displayTextRange": [
    0,
    276
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "omarsar0",
    "url": "https://x.com/omarsar0",
    "twitterUrl": "https://twitter.com/omarsar0",
    "id": "3448284313",
    "name": "elvis",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/939313677647282181/vZjFWtAn_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/3448284313/1565974901",
    "description": "",
    "location": "DAIR.AI Academy",
    "followers": 293970,
    "following": 789,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Fri Sep 04 12:59:26 +0000 2015",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 35155,
    "hasCustomTimelines": true,
    "isTranslator": true,
    "mediaCount": 4559,
    "statusesCount": 17503,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2032822511709147640"
    ],
    "profile_bio": {
      "description": "Building @dair_ai • Prev: Meta AI, Elastic, PhD • New AI learning portal: https://t.co/1e8RZKs4uX",
      "entities": {
        "description": {
          "hashtags": [],
          "symbols": [],
          "urls": [
            {
              "display_url": "academy.dair.ai",
              "expanded_url": "https://academy.dair.ai/",
              "indices": [
                74,
                97
              ],
              "url": "https://t.co/1e8RZKs4uX"
            }
          ],
          "user_mentions": [
            {
              "id_str": "0",
              "indices": [
                9,
                17
              ],
              "name": "",
              "screen_name": "dair_ai"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/XQto5ypSIk"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/p36Vy91gnD",
        "expanded_url": "https://twitter.com/omarsar0/status/2032928526022881399/photo/1",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2032928522403246080",
        "indices": [
          277,
          300
        ],
        "media_key": "3_2032928522403246080",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARw2abhfmvAACgACHDZpuTdaMHcAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHDZpuF+a8AAKAAIcNmm5N1owdwAA",
            "media_key": "3_2032928522403246080"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HDZpuF-a8AAW04c.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 903,
              "w": 1612,
              "x": 0,
              "y": 0
            },
            {
              "h": 1612,
              "w": 1612,
              "x": 0,
              "y": 0
            },
            {
              "h": 1784,
              "w": 1565,
              "x": 47,
              "y": 0
            },
            {
              "h": 1784,
              "w": 892,
              "x": 579,
              "y": 0
            },
            {
              "h": 1784,
              "w": 1612,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1784,
          "width": 1612
        },
        "sizes": {
          "large": {
            "h": 1784,
            "w": 1612
          }
        },
        "type": "photo",
        "url": "https://t.co/p36Vy91gnD"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2603.12056",
        "expanded_url": "https://arxiv.org/abs/2603.12056",
        "indices": [
          1327,
          1350
        ],
        "url": "https://t.co/YXrHcJ6Zim"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          1403,
          1426
        ],
        "url": "https://t.co/1e8RZKs4uX"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}