🐦 Twitter Post Details

Viewing enriched Twitter post

@dair_ai

// Coding Agents are Effective Long-Context Processors // We are just touching the surface of what's possible with coding agents. LLMs struggle with long contexts, even the ones that support massive context windows. It turns out coding agents already know how to solve this; you just need to reframe the problem. This work places massive text corpora into directory structures and lets off-the-shelf coding agents (Codex, Claude Code) navigate them with terminal commands and Python scripts. This is great, as you are not feeding massive text directly into a model’s context window or relying on semantic retrieval. Results: - On BrowseComp-Plus (750M tokens), this approach scores 88.5% vs 80% best published. - On Oolong-Real (385K tokens), 33.7% vs 24.1%, a 56% relative improvement. - GPT-5 full-context baseline only manages 20% on BrowseComp-Plus. Works up to 3 trillion tokens. Instead of scaling context windows or building retrieval pipelines, coding agents that already know how to navigate file systems can process virtually unlimited context. The agents autonomously develop task-specific strategies: writing scripts, iterative query refinement, and programmatic aggregation. Paper: https://t.co/kiCFtlixMf Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c

Media 1
Media 2

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2038635382989005015/media_0.jpg",
      "filename": "media_0.jpg"
    },
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2038635382989005015/media_1.png",
      "filename": "media_1.png"
    }
  ],
  "processed_at": "2026-03-30T15:16:49.929678",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2038635382989005015",
  "url": "https://x.com/dair_ai/status/2038635382989005015",
  "twitterUrl": "https://twitter.com/dair_ai/status/2038635382989005015",
  "text": "// Coding Agents are Effective Long-Context Processors //\n\nWe are just touching the surface of what's possible with coding agents.\n\nLLMs struggle with long contexts, even the ones that support massive context windows.\n\nIt turns out coding agents already know how to solve this; you just need to reframe the problem.\n\nThis work places massive text corpora into directory structures and lets off-the-shelf coding agents (Codex, Claude Code) navigate them with terminal commands and Python scripts.\n\nThis is great, as you are not feeding massive text directly into a model’s context window or relying on semantic retrieval.\n\nResults:\n\n- On BrowseComp-Plus (750M tokens), this approach scores 88.5% vs 80% best published.\n\n- On Oolong-Real (385K tokens), 33.7% vs 24.1%, a 56% relative improvement.\n\n- GPT-5 full-context baseline only manages 20% on BrowseComp-Plus.\n\nWorks up to 3 trillion tokens.\n\nInstead of scaling context windows or building retrieval pipelines, coding agents that already know how to navigate file systems can process virtually unlimited context.\n\nThe agents autonomously develop task-specific strategies: writing scripts, iterative query refinement, and programmatic aggregation.\n\nPaper: https://t.co/kiCFtlixMf\n\nLearn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c",
  "source": "Twitter for iPhone",
  "retweetCount": 0,
  "replyCount": 0,
  "likeCount": 4,
  "quoteCount": 0,
  "viewCount": 175,
  "createdAt": "Mon Mar 30 15:12:05 +0000 2026",
  "lang": "en",
  "bookmarkCount": 5,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2038635382989005015",
  "displayTextRange": [
    0,
    277
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "dair_ai",
    "url": "https://x.com/dair_ai",
    "twitterUrl": "https://twitter.com/dair_ai",
    "id": "889050642903293953",
    "name": "DAIR.AI",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1643277398522187778/31dedbLo_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/889050642903293953/1773242460",
    "description": "",
    "location": "",
    "followers": 93601,
    "following": 1,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sun Jul 23 09:12:45 +0000 2017",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 4290,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 181,
    "statusesCount": 3050,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2038262704486400218"
    ],
    "profile_bio": {
      "description": "Democratizing AI research, education, and technologies. New AI learning portal: https://t.co/LRnpZN7L4c",
      "entities": {
        "description": {
          "hashtags": [],
          "symbols": [],
          "urls": [
            {
              "display_url": "academy.dair.ai",
              "expanded_url": "https://academy.dair.ai/",
              "indices": [
                80,
                103
              ],
              "url": "https://t.co/LRnpZN7L4c"
            }
          ],
          "user_mentions": []
        },
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/lkqPZtMU5s"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/gIPMLfzish",
        "expanded_url": "https://twitter.com/dair_ai/status/2038635382989005015/photo/1",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {
            "faces": [
              {
                "h": 595,
                "w": 595,
                "x": 43,
                "y": 564
              }
            ]
          },
          "orig": {
            "faces": [
              {
                "h": 595,
                "w": 595,
                "x": 43,
                "y": 564
              }
            ]
          }
        },
        "id_str": "2038635379465891840",
        "indices": [
          278,
          301
        ],
        "media_key": "3_2038635379465891840",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARxKsBOTm7AACgACHEqwFGWaINcAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHEqwE5ObsAAKAAIcSrAUZZog1wAA",
            "media_key": "3_2038635379465891840"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HEqwE5ObsAAjqig.jpg",
        "original_info": {
          "focus_rects": [
            {
              "h": 778,
              "w": 1390,
              "x": 0,
              "y": 0
            },
            {
              "h": 1390,
              "w": 1390,
              "x": 0,
              "y": 0
            },
            {
              "h": 1536,
              "w": 1347,
              "x": 0,
              "y": 0
            },
            {
              "h": 1536,
              "w": 768,
              "x": 114,
              "y": 0
            },
            {
              "h": 1536,
              "w": 1390,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1536,
          "width": 1390
        },
        "sizes": {
          "large": {
            "h": 1536,
            "w": 1390
          }
        },
        "type": "photo",
        "url": "https://t.co/gIPMLfzish"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2603.20432",
        "expanded_url": "https://arxiv.org/abs/2603.20432",
        "indices": [
          1208,
          1231
        ],
        "url": "https://t.co/kiCFtlixMf"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          1284,
          1307
        ],
        "url": "https://t.co/LRnpZN7L4c"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}