🐦 Twitter Post Details

Viewing enriched Twitter post

@omarsar0

// Automating SKILL.md Generation // Increasingly, mining sessions is one of the best ways to improve your agents. OpenAI released something similar yesterday that lets Codex package skills from interactions. (bookmark it) This paper explains a related approach. They run a three-stage pipeline that segments GUI trajectories, clusters them into candidate skills, and trains a skill-aware policy. The clusters are genuinely readable, with five of eight hitting 0.95 or higher purity against ground-truth workflow labels. But readability does not transfer. GRPO lifts skill-step accuracy only from 18.5% to 20.5%, leaves BrowseComp+ flat, and loses to trivial frequency priors. The authors name the three culprits: a weak boundary detector, an orderless segment representation, and an offline reward model. Paper: https://t.co/Du48U4xNwX Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX

Media 1
View on Twitter

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2067986774241251433/media_0.png",
      "filename": "media_0.png"
    }
  ],
  "processed_at": "2026-06-19T15:04:57.671691",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2067986774241251433",
  "url": "https://x.com/omarsar0/status/2067986774241251433",
  "twitterUrl": "https://twitter.com/omarsar0/status/2067986774241251433",
  "text": "// Automating SKILL.md Generation //\n\nIncreasingly, mining sessions is one of the best ways to improve your agents.\n\nOpenAI released something similar yesterday that lets Codex package skills from interactions.\n\n(bookmark it)\n\nThis paper explains a related approach.\n\nThey run a three-stage pipeline that segments GUI trajectories, clusters them into candidate skills, and trains a skill-aware policy.\n\nThe clusters are genuinely readable, with five of eight hitting 0.95 or higher purity against ground-truth workflow labels.\n\nBut readability does not transfer. GRPO lifts skill-step accuracy only from 18.5% to 20.5%, leaves BrowseComp+ flat, and loses to trivial frequency priors.\n\nThe authors name the three culprits: a weak boundary detector, an orderless segment representation, and an offline reward model.\n\nPaper: https://t.co/Du48U4xNwX\n\nLearn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX",
  "source": "Twitter for iPhone",
  "retweetCount": 0,
  "replyCount": 0,
  "likeCount": 1,
  "quoteCount": 0,
  "viewCount": 20,
  "createdAt": "Fri Jun 19 15:04:02 +0000 2026",
  "lang": "en",
  "bookmarkCount": 1,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2067986774241251433",
  "displayTextRange": [
    0,
    278
  ],
  "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": 307827,
    "following": 870,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Fri Sep 04 12:59:26 +0000 2015",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 36809,
    "hasCustomTimelines": true,
    "isTranslator": true,
    "mediaCount": 4727,
    "statusesCount": 18335,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2067952726282031411"
    ],
    "profile_bio": {
      "description": "Building self-improving AI @dair_ai • Prev: Meta AI | PhD • Learn about AI Agents for FREE here: https://t.co/P5SA9u54xO",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "academy.dair.ai/courses/elemen…",
              "expanded_url": "https://academy.dair.ai/courses/elements-of-ai-agents",
              "indices": [
                97,
                120
              ],
              "url": "https://t.co/P5SA9u54xO"
            }
          ],
          "user_mentions": [
            {
              "id_str": "",
              "indices": [
                27,
                35
              ],
              "name": "",
              "screen_name": "dair_ai"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/XQto5ypkSM"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/HyVvbtF1MN",
        "expanded_url": "https://twitter.com/omarsar0/status/2067986774241251433/photo/1",
        "ext_master_playlist_only": [],
        "ext_media_availability": {
          "status": "Available"
        },
        "ext_playlists": [],
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2067986769543667712",
        "indices": [
          279,
          302
        ],
        "media_key": "3_2067986769543667712",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARyy9wKe26AACgACHLL3A7bbEGkAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHLL3Ap7boAAKAAIcsvcDttsQaQAA",
            "media_key": "3_2067986769543667712"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HLL3Ap7boAAMAyV.png",
        "original_info": {
          "focus_rects": [
            {
              "h": 896,
              "w": 1600,
              "x": 0,
              "y": 0
            },
            {
              "h": 1600,
              "w": 1600,
              "x": 0,
              "y": 0
            },
            {
              "h": 1770,
              "w": 1553,
              "x": 24,
              "y": 0
            },
            {
              "h": 1770,
              "w": 885,
              "x": 358,
              "y": 0
            },
            {
              "h": 1770,
              "w": 1600,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1770,
          "width": 1600
        },
        "sizes": {
          "large": {
            "h": 1770,
            "w": 1600
          }
        },
        "type": "photo",
        "url": "https://t.co/HyVvbtF1MN"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2606.20363",
        "expanded_url": "https://arxiv.org/abs/2606.20363",
        "indices": [
          822,
          845
        ],
        "url": "https://t.co/Du48U4xNwX"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          898,
          921
        ],
        "url": "https://t.co/1e8RZKs4uX"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "communityInfo": null,
  "article": null
}