{"slug": "codex-is-eating-my-mac-s-disk-space", "title": "Codex is EATING my Mac's disk space", "summary": "A GitHub issue filed against OpenAI's Codex repository reports that the Codex Desktop app for macOS (version 26.527.60818, CFBundleVersion 3437) continuously writes paired .dmp and _sidecar.json Crashpad files to ~/Library/Application Support/com.openai.codex/web/Crashpad/pending, reaching 4.9GB across 54,504 files — 27,252 .dmp and 27,252 _sidecar.json — in a single day. The reporter measured the file count climbing from 53,444 to 54,064 to 54,504 during the investigation while the app stayed usable, and observed two browser_crashpad_handler processes running under PPID=1 since May 30, 2026 sharing the same Crashpad database. The reporter says the dumps may contain memory, local paths or private data and is withholding minidumps publicly pending a safe upload path from OpenAI maintainers.", "body_md": "## Repository navigation\n\n# Codex Desktop continuously generates Crashpad pending dumps, growing without any limit: +5GB per day at least #25921\n\n[appIssues related to the Codex desktop app](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22app%22)Issues related to the Codex desktop app\n\n[bugSomething isn't working](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22bug%22)Something isn't working\n\n[performance](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22performance%22)\n\n## Description\n\n## Summary\n\nCodex Desktop is continuously adding paired `.dmp` and `_sidecar.json` files under:\n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad/pending\n```\n\nOn my machine this directory grew to `4.9G` and `54,504` files in one day. The count continued increasing during investigation, so this appears to be active Crashpad dump generation rather than stale one-time leftovers.\n\nThe Codex Desktop app is still functional while this is happening. I can use the app, inspect the filesystem, and interact with GitHub from the same running session.\n\nSo the user-visible problem is not that Codex is fully crashing or unable to launch. The issue is that Crashpad is continuously generating pending dump/sidecar pairs in the background while the app remains usable, causing unbounded local disk growth.\n\nI am not attaching minidumps publicly because they may contain memory, local paths, or private data. I can share the newest `.dmp` privately if OpenAI maintainers provide a safe upload path.\n\n## Environment\n\n- Product: Codex Desktop app on macOS\n- Codex app version: `26.527.60818`\n- `CFBundleVersion` :`3437`\n- Crashpad/Electron annotation observed from live process: `prod=Codex_Mac` ,`ver=148.0.7778.179`\n- macOS: `26.3.1 (25D771280a)`\n- Architecture: `arm64`\n\n## Observed evidence\n\nCrashpad pending directory:\n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad/pending\n```\n\nLatest measurements:\n\n```\n4.9G    ~/Library/Application Support/com.openai.codex/web/Crashpad/pending\n54,504 files total\n27,252 .dmp files\n27,252 _sidecar.json files\n```\n\nSampled pending file time range:\n\n```\noldest sampled pending files: 2026-06-02 13:31:21 CEST\nlatest sampled pending files: 2026-06-02 21:08:24 CEST\n```\n\nDuring the investigation the count increased from `53,444` to `54,064`, then to `54,504`, which indicates the queue was still growing.\n\nI also observed two Codex `browser_crashpad_handler` processes running under `PPID=1` since `Sat May 30 11:47:50 2026`, both using the same Crashpad database:\n\n```\n/Applications/Codex.app/.../Helpers/browser_crashpad_handler \\\n  --database=~/Library/Application Support/com.openai.codex/web/Crashpad \\\n  --annotation=plat=OS X \\\n  --annotation=prod=Codex_Mac \\\n  --annotation=ver=148.0.7778.179\n```\n\nThe local app remained usable enough to inspect/report this, so the visible issue is not just a normal foreground app crash. It looks like Crashpad or a monitored helper process is repeatedly generating pending dumps without bounding, upload/processing cleanup, or rotation.\n\n## Expected behavior\n\nCrash reports should be bounded, uploaded/processed/rotated, or not continuously generated while the app remains usable.\n\n## Actual behavior\n\nPending Crashpad reports accumulate rapidly and consume multiple GB of local disk space.\n\n## Commands used to collect evidence\n\n```\ndu -sh \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\"\n\nfind \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\" -maxdepth 1 -type f | wc -l\n\nfind \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\" -maxdepth 1 -type f \\\n  | sed 's/.*\\.//' | sort | uniq -c | sort -nr\n\n/usr/libexec/PlistBuddy -c 'Print :CFBundleShortVersionString' /Applications/Codex.app/Contents/Info.plist\n/usr/libexec/PlistBuddy -c 'Print :CFBundleVersion' /Applications/Codex.app/Contents/Info.plist\n\nsw_vers\nuname -m\n\nps ax -o pid,ppid,lstart,command \\\n  | grep -E 'browser_crashpad_handler|com.openai.codex/web/Crashpad' \\\n  | grep -v grep\n```\n\n## Related issues\n\nThis may be related to other Codex Desktop helper/process cleanup reports:\n\n- [Codex App update left old crashpad helper; relaunch could no-op until clean restart #25259](https://github.com/openai/codex/issues/25259) : stale Crashpad helper after update/relaunch\n- [Codex App UI freezes: worker crashes, massive git/worktree errors, and concurrent orphaned app-server  processes #11090](https://github.com/openai/codex/issues/11090) : orphaned Codex processes and UI freezes\n- [Can't open app in Tahoe 26.2 (25C56) #10627](https://github.com/openai/codex/issues/10627) : prior Crashpad/macOS startup reports\n\n## Activity\n\n1. added[bugSomething isn't working](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22bug%22) Something isn't working[appIssues related to the Codex desktop app](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22app%22) Issues related to the Codex desktop appon Jun 2, 2026\n2. ### Jolg42 commentedon Jun 2, 2026 AuthorMore actionsUpdate: I cleared the files inside the Crashpad `pending` directory after filing this issue.Result immediately after cleanup: \n\n```\nbefore cleanup: 4.9G, 54,504 files\ncleanup action: deleted files under ~/Library/Application Support/com.openai.codex/web/Crashpad/pending\nimmediate post-cleanup state: 6.4M, 70 files\npost-cleanup split: 35 .dmp files, 35 _sidecar.json files\n```\n\n So the bulk of the disk usage was recoverable by deleting pending reports, but new paired dump/sidecar files appeared immediately afterward. That supports the earlier observation that Crashpad dump generation is still active, not just a stale historical backlog. I did not attach any `.dmp` files publicly.\n3. I am seeing the same issue on macOS with a newer Codex Desktop build, with a bit more diagnostic detail. ## Environment\n  - Product: Codex Desktop app on macOS\n  - Codex app version: `26.602.71036`\n  - `CFBundleVersion` :`3685`\n  - Framework/Crashpad annotation observed in dumps: `ver=149.0.7827.54`\n  - Architecture: arm64\n  - Crashpad database path:\n \n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad\n```\n\n ## Impact observedThe pending directory grew to about `61G` :\n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad/pending\n```\n\n Counts from the pending directory before cleanup: \n\n```\nfiles: 738,864\n.dmp files: 369,431\n_sidecar.json files: 369,433\nbytes: 63,233,668,274\nGiB: 58.89\nrecent_24h files: 98,218\nrecent_7d files: 573,480\noldest: 2026-06-01 10:52:52\nnewest: 2026-06-09 10:00:26\n```\n\n After cleanup/reinstall/restart, the directory continued growing at roughly one crash report per second: \n\n```\n10:28:46 files=142 MiB=11.63\n10:28:51 files=152 MiB=12.44\n10:28:56 files=162 MiB=13.26\n10:29:01 files=172 MiB=14.08\n10:29:06 files=182 MiB=14.89\n```\n\n ## Key diagnostic detailThe newest `.dmp` files contain this FATAL message from`browser_crashpad_handler` itself:\n\n```\nbrowser_crashpad_handler\nCodex_Mac\ncrashpad-handler\n[0609/103121.674299:FATAL:third_party/crashpad/crashpad/handler/mac/exception_handler_server.cc:76]\nCheck failed: kr == KERN_SUCCESS.\nmach_port_request_notification: (os/kern) invalid capability (20)\nexecutable_path=/Applications/Codex.app/Contents/Frameworks/Codex Framework.framework/Versions/149.0.7827.54/Helpers/browser_crashpad_handler\n--database=/Users/<user>/Library/Application Support/com.openai.codex/web/Crashpad\n--annotation=prod=Codex_Mac\n```\n\n So in this case it looks like `browser_crashpad_handler` is crashing/restarting in a loop and writing paired`.dmp` /`_sidecar.json` files for itself.System diagnostic logs also show repeated `browser_crashpad_handler` crashes with:\n\n```\nprocName: browser_crashpad_handler\nprocPath: /Applications/Codex.app/Contents/Frameworks/Codex Framework.framework/Versions/149.0.7827.54/Helpers/browser_crashpad_handler\ncoalitionName: com.openai.codex\nexception: EXC_BREAKPOINT / SIGTRAP\n```\n\n ## Related behaviorThe in-app browser / Browser Use had previously become unavailable on this machine. During investigation, Codex logs repeatedly showed Browser Use / IAB lifecycle entries, and `browser_use_availability_resolved available=true` shortly before Crashpad kept generating dumps. That may or may not be the trigger, but the issue appears tied to the Codex Desktop browser/Chromium/Crashpad path.A reinstall did not fix it. Disabling `browser@openai-bundled` , disabling`chrome@openai-bundled` , and setting`BROWSER_USE_AVAILABLE_BACKENDS = \"\"` in`~/.codex/config.toml` also did not stop the dump growth after restart.## Temporary workaround that stopped disk growthBecause macOS would not allow modifying `/Applications/Codex.app/.../browser_crashpad_handler` even with`sudo` (`Operation not permitted` ), I worked around it by making the user Crashpad database read-only and immutable:\n\n```\ncrash=\"$HOME/Library/Application Support/com.openai.codex/web/Crashpad\"\npkill -f \"browser_crashpad_handler.*com.openai.codex/web/Crashpad\" 2>/dev/null || true\nchflags -R nouchg \"$crash\" 2>/dev/null || true\nchmod -R u+w \"$crash\" 2>/dev/null || true\nrm -rf \"$crash/pending\" \"$crash/new\" \"$crash/completed\" \"$crash/attachments\"\nmkdir -p \"$crash/pending\" \"$crash/new\" \"$crash/completed\" \"$crash/attachments\"\nchmod -R 555 \"$crash\"\nchflags -R uchg \"$crash\"\n```\n\n After that, pending stayed at zero for the observed window: \n\n```\n10:40:12 files=0 MiB=0.0\n10:40:17 files=0 MiB=0.0\n10:40:22 files=0 MiB=0.0\n10:40:27 files=0 MiB=0.0\n10:40:32 files=0 MiB=0.0\n10:40:37 files=0 MiB=0.0\n10:40:42 files=0 MiB=0.0\n10:40:47 files=0 MiB=0.0\n```\n\n This is only a disk-growth workaround, not a real fix, since it effectively prevents Codex from writing Crashpad reports. Reacted by ai-poet\n4. I am too, just cleaned up 22G today\n5. changed the title~~[-]Codex Desktop continuously generates Crashpad pending dumps, growing to ~5GB[/-]~~ [+]Codex Desktop continuously generates Crashpad pending dumps, growing without any limit: +5GB per day at least[/+]on Jun 9, 2026\n6. same issue on version 26.601.10930. cc [@dnakov](https://github.com/dnakov)\n7. I think this is fixed with later macos versions. at least in 26.5 it is Reacted by Jonathan Chang\n8. I ran into this exact same issue. Oh my god, no wonder my Mac's storage keeps getting full. Is Codex Desktop just a mountain of garbage code OpenAI built via vibe coding? Reacted by jordanade\n9. Both Claude and Codex have ZERO RESPECT for user's machines. Holy hell. Reacted by louie-lou and James Power\n10. every json just 2 bytes+ 150KB, but the more serious thing behind is that, there are crash reports continously generating in `/System/Library/Caches/com.apple.coresymbolicationd/data` use this script to check your macOS. \n\n``` bash\n#!/usr/bin/env bash\nset -uo pipefail\n\n# recent-files.sh\n#\n# Usage:\n#   ./recent-files.sh [directory] [hours]\n#\n# Examples:\n#   ./recent-files.sh\n#   ./recent-files.sh / 24\n#   sudo ./recent-files.sh / 168\n#\n# Environment:\n#   PAGE_SIZE=500 ./recent-files.sh / 24\n#   FORCE_SCAN=1 ./recent-files.sh / 24   # macOS: bypass Spotlight and scan\n\nDIR=\"${1:-.}\"\nHOURS=\"${2:-24}\"\nPAGE_SIZE=\"${PAGE_SIZE:-1000}\"\nFORCE_SCAN=\"${FORCE_SCAN:-0}\"\n\nif [[ ! -d \"$DIR\" ]]; then\n  echo \"Error: '$DIR' is not a directory\"\n  exit 1\nfi\n\nif ! [[ \"$HOURS\" =~ ^[0-9]+$ ]]; then\n  echo \"Error: hours must be a positive integer\"\n  exit 1\nfi\n\nif ! [[ \"$PAGE_SIZE\" =~ ^[0-9]+$ ]]; then\n  echo \"Error: PAGE_SIZE must be a positive integer\"\n  exit 1\nfi\n\nTMP=\"$(mktemp \"${TMPDIR:-/tmp}/recent-files.XXXXXX\")\"\ntrap 'rm -f \"$TMP\"' EXIT\n\nPAGE=0\n\nis_gnu_find() {\n  find --version >/dev/null 2>&1\n}\n\nis_macos() {\n  [[ \"$(uname -s)\" == \"Darwin\" ]]\n}\n\nindex_linux_gnu_find() {\n  local minutes=$((HOURS * 60))\n\n  # Output fields:\n  # sort_epoch | size | added_date | modified_date | file_path\n  find \"$DIR\" \\\n    -type f \\\n    -mmin \"-$minutes\" \\\n    -printf '%T@\\t%s\\t%w\\t%TY-%Tm-%Td %TH:%TM:%TS\\t%p\\n' \\\n    2>/dev/null \\\n    | sort -nr > \"$TMP\"\n}\n\nindex_macos_spotlight() {\n  local cutoff_iso\n  cutoff_iso=\"$(date -v-\"${HOURS}\"H '+%Y-%m-%d %H:%M:%S')\"\n\n  # Spotlight is indexed and very fast, but may exclude some system/protected areas.\n  # kMDItemFSContentChangeDate = modified/change date.\n  mdfind -onlyin \"$DIR\" \\\n    \"kMDItemFSContentChangeDate >= \\$time.iso($cutoff_iso)\" \\\n    2>/dev/null \\\n    | while IFS= read -r file; do\n        [[ -f \"$file\" ]] || continue\n\n        mtime_epoch=\"$(stat -f '%m' \"$file\" 2>/dev/null)\" || continue\n        size=\"$(stat -f '%z' \"$file\" 2>/dev/null || echo \"-\")\"\n        added=\"$(stat -f '%SB' -t '%Y-%m-%d %H:%M:%S' \"$file\" 2>/dev/null || echo \"-\")\"\n        modified=\"$(stat -f '%Sm' -t '%Y-%m-%d %H:%M:%S' \"$file\" 2>/dev/null || echo \"-\")\"\n\n        printf '%s\\t%s\\t%s\\t%s\\t%s\\n' \\\n          \"$mtime_epoch\" \"$size\" \"$added\" \"$modified\" \"$file\"\n      done \\\n      | sort -nr > \"$TMP\"\n}\n\nindex_macos_full_scan() {\n  local now cutoff\n  now=\"$(date +%s)\"\n  cutoff=$((now - HOURS * 3600))\n\n  # Slower than Spotlight, but searches actual filesystem paths recursively.\n  find \"$DIR\" -type f -print0 2>/dev/null \\\n    | while IFS= read -r -d '' file; do\n        mtime_epoch=\"$(stat -f '%m' \"$file\" 2>/dev/null)\" || continue\n        (( mtime_epoch >= cutoff )) || continue\n\n        size=\"$(stat -f '%z' \"$file\" 2>/dev/null || echo \"-\")\"\n        added=\"$(stat -f '%SB' -t '%Y-%m-%d %H:%M:%S' \"$file\" 2>/dev/null || echo \"-\")\"\n        modified=\"$(stat -f '%Sm' -t '%Y-%m-%d %H:%M:%S' \"$file\" 2>/dev/null || echo \"-\")\"\n\n        printf '%s\\t%s\\t%s\\t%s\\t%s\\n' \\\n          \"$mtime_epoch\" \"$size\" \"$added\" \"$modified\" \"$file\"\n      done \\\n      | sort -nr > \"$TMP\"\n}\n\nbuild_index() {\n  : > \"$TMP\"\n\n  echo \"Indexing recent files...\"\n  echo \"Directory: $DIR\"\n  echo \"Recent window: last $HOURS hour(s)\"\n  echo\n\n  if is_gnu_find; then\n    index_linux_gnu_find\n  elif is_macos && command -v mdfind >/dev/null 2>&1 && [[ \"$FORCE_SCAN\" != \"1\" ]]; then\n    index_macos_spotlight\n  else\n    index_macos_full_scan\n  fi\n\n  local count\n  count=\"$(wc -l < \"$TMP\" | tr -d ' ')\"\n\n  if [[ \"$count\" == \"0\" ]]; then\n    echo \"No recent files found.\"\n    echo\n    echo \"Tips:\"\n    echo \"  - Try a larger window: ./recent-files.sh '$DIR' 168\"\n    echo \"  - Search from root: sudo ./recent-files.sh / 24\"\n    echo \"  - On macOS, force real scan: FORCE_SCAN=1 ./recent-files.sh '$DIR' 24\"\n    exit 0\n  fi\n}\n\nprint_page() {\n  local start end total\n\n  total=\"$(wc -l < \"$TMP\" | tr -d ' ')\"\n  start=$((PAGE * PAGE_SIZE + 1))\n  end=$((start + PAGE_SIZE - 1))\n\n  clear\n  echo \"Directory: $DIR\"\n  echo \"Recent window: last $HOURS hour(s)\"\n  echo \"Total files found: $total\"\n  echo \"Page: $((PAGE + 1))\"\n  echo \"Showing: $start-$end\"\n  echo\n\n  printf \"%-12s | %-19s | %-19s | %s\\n\" \\\n    \"Size(bytes)\" \"Added date\" \"Modified date\" \"File\"\n  printf \"%-12s-+-%-19s-+-%-19s-+-%s\\n\" \\\n    \"------------\" \"-------------------\" \"-------------------\" \"------------------------------\"\n\n  sed -n \"${start},${end}p\" \"$TMP\" \\\n    | awk -F '\\t' '{\n        # $1 = sort epoch\n        # $2 = size\n        # $3 = added date\n        # $4 = modified date\n        # $5+ = file path, in case path somehow contains tabs\n        file=$5\n        for (i=6; i<=NF; i++) file=file \"\\t\" $i\n\n        sub(/\\..*/, \"\", $3)\n        sub(/\\..*/, \"\", $4)\n\n        printf \"%-12s | %-19s | %-19s | %s\\n\", $2, $3, $4, file\n      }'\n}\n\nbuild_index\n\nwhile true; do\n  print_page\n\n  echo\n  echo \"[n] next page | [p] previous page | [r] rebuild | [q] quit\"\n  read -r -p \"Choice: \" choice\n\n  case \"$choice\" in\n    n|N)\n      PAGE=$((PAGE + 1))\n      ;;\n    p|P)\n      (( PAGE > 0 )) && PAGE=$((PAGE - 1))\n      ;;\n    r|R)\n      PAGE=0\n      build_index\n      ;;\n    q|Q)\n      exit 0\n      ;;\n    *)\n      echo \"Unknown choice.\"\n      sleep 1\n      ;;\n  esac\ndone\n```\n\n \n script to view and clean the crash reports \n\n``` bash\n#!/bin/bash\n\nTARGET_DIR=\"/System/Library/Caches/com.apple.coresymbolicationd\"\nTARGET_DATA=\"$TARGET_DIR/data\"\n\nshow_menu() {\n  echo\n  echo \"macOS coresymbolicationd cache tool\"\n  echo \"Target: $TARGET_DIR\"\n  echo\n  echo \"1) View / check files inside com.apple.coresymbolicationd\"\n  echo \"2) Clean up com.apple.coresymbolicationd cache\"\n  echo \"q) Quit\"\n  echo\n}\n\nview_cache() {\n  echo\n  echo \"Checking: $TARGET_DIR\"\n  echo\n\n  if [ ! -d \"$TARGET_DIR\" ]; then\n    echo \"Directory not found: $TARGET_DIR\"\n    return\n  fi\n\n  echo \"Total size:\"\n  sudo du -sh \"$TARGET_DIR\" 2>/dev/null || echo \"Unable to read size. Try running with sudo.\"\n  echo\n\n  echo \"Files:\"\n  sudo ls -lah \"$TARGET_DIR\"\n  echo\n\n  if [ -f \"$TARGET_DATA\" ]; then\n    echo \"Main data file:\"\n    sudo ls -lh \"$TARGET_DATA\"\n    echo\n\n    echo \"Approx size in GB:\"\n    sudo du -g \"$TARGET_DATA\" 2>/dev/null\n  else\n    echo \"No data file found at: $TARGET_DATA\"\n  fi\n\n  echo\n  echo \"Recently modified items:\"\n  sudo find \"$TARGET_DIR\" -maxdepth 2 -type f -print0 2>/dev/null \\\n    | sudo xargs -0 ls -lt 2>/dev/null \\\n    | head -20\n}\n\nclean_cache() {\n  echo\n  echo \"This will delete files inside:\"\n  echo \"$TARGET_DIR\"\n  echo\n  echo \"It will NOT delete the folder itself.\"\n  echo \"macOS should recreate the cache when needed.\"\n  echo\n\n  read -r -p \"Type CLEAN to continue: \" confirm\n\n  if [ \"$confirm\" != \"CLEAN\" ]; then\n    echo \"Cancelled.\"\n    return\n  fi\n\n  echo\n  echo \"Stopping coresymbolicationd if running...\"\n  sudo killall coresymbolicationd 2>/dev/null || true\n\n  echo \"Deleting cache contents...\"\n  if [ -d \"$TARGET_DIR\" ]; then\n    sudo find \"$TARGET_DIR\" -mindepth 1 -maxdepth 1 -exec rm -rf {} +\n    echo \"Cleanup complete.\"\n  else\n    echo \"Directory not found: $TARGET_DIR\"\n  fi\n\n  echo\n  echo \"Current size:\"\n  sudo du -sh \"$TARGET_DIR\" 2>/dev/null || echo \"Cache directory may not exist until macOS recreates it.\"\n\n  echo\n  echo \"Tip: reboot if the storage number in System Settings does not update immediately.\"\n}\n\nwhile true; do\n  show_menu\n  read -r -p \"Choose an option: \" choice\n\n  case \"$choice\" in\n    1)\n      view_cache\n      ;;\n    2)\n      clean_cache\n      ;;\n    q|Q)\n      echo \"Bye.\"\n      exit 0\n      ;;\n    *)\n      echo \"Invalid option.\"\n      ;;\n  esac\ndone\n```\n\n \n after deleting, try delete this as well to reset Codex. \n\n```\n~/Library/Saved Application State/com.openai.codex.savedState\n```\n\n11. I hit what appears to be the same Crashpad pending-dump growth issue, but at a much larger scale. ## Environment\n  - Product: Codex Desktop app on macOS\n  - Codex app version: `26.611.62324`\n  - `CFBundleVersion` :`4028`\n  - macOS: `15.6 (24G84)`\n  - Architecture: `arm64`\n  - Current helper/Crashpad annotation observed from live process path: Electron/Chromium helper `149.0.7827.115`\n ## Observed disk usageThe large growth was under the legacy app-data path: \n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad/pending\n```\n\n Before cleanup, this directory contained: \n\n```\n1,786,086 total entries\n893,043 .dmp files\n893,043 _sidecar.json files\n147,440,897,945 bytes total\n137.32 GiB total\n137.31 GiB from .dmp files\n```\n\n File time range: \n\n```\noldest: 2026-06-01 11:55:53\nnewest: 2026-06-18 14:16:58\n```\n\n Daily distribution included multiple days above 13 GiB: \n\n```\n2026-06-06  13.17 GiB\n2026-06-07  13.18 GiB\n2026-06-08  13.18 GiB\n2026-06-12  13.14 GiB\n2026-06-13  13.17 GiB\n2026-06-14  13.17 GiB\n2026-06-15  11.66 GiB\n2026-06-18   7.82 GiB\n```\n\n Largest observed dump files were still relatively small: \n\n```\n782,272 bytes\n739,984 bytes\n705,456 bytes\n```\n\n So the disk growth was caused by hundreds of thousands of small crash dumps, not a few large files. ## Cleanup resultAfter deleting pending files, the legacy path is now empty: \n\n```\n~/Library/Application Support/com.openai.codex/web/Crashpad/pending\ncount=0\ntotal_mib=0.00\n```\n\n The currently running Codex app appears to use a different app-data path: \n\n```\n~/Library/Application Support/Codex/Crashpad\n```\n\n Current size there is small: \n\n```\n~/Library/Application Support/Codex                 223M\n~/Library/Application Support/Codex/Crashpad         20K\n~/Library/Application Support/Codex/Crashpad/pending 16K\npending count=4\n```\n\n That path difference may be relevant: the unbounded growth happened under `com.openai.codex/web/Crashpad` , while the current running app uses`Codex/Crashpad` .I am not attaching `.dmp` files publicly because they may contain private memory, paths, or local state. I can collect additional metadata or share a dump through a safe maintainer-provided upload path if useful.\n12. For anyone still trying to distinguish \"old backlog\" from \"active dump loop\", I would measure both Codex app-data paths separately before deleting broader app state: \n\n```\npaths=(\n  \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad\"\n  \"$HOME/Library/Application Support/Codex/Crashpad\"\n)\n\nfor crash in \"${paths[@]}\"; do\n  echo \"== $crash ==\"\n  [ -d \"$crash\" ] || { echo \"missing\"; continue; }\n  du -sh \"$crash\" \"$crash/pending\" 2>/dev/null\n  find \"$crash/pending\" -maxdepth 1 -type f 2>/dev/null | wc -l\n  find \"$crash/pending\" -maxdepth 1 -type f -name '*.dmp' 2>/dev/null | wc -l\n  find \"$crash/pending\" -maxdepth 1 -type f -name '*_sidecar.json' 2>/dev/null | wc -l\n  find \"$crash/pending\" -maxdepth 1 -type f -print0 2>/dev/null \\\n    | xargs -0 stat -f '%Sm %z %N' -t '%Y-%m-%d %H:%M:%S' 2>/dev/null \\\n    | sort | sed -n '1p;$p'\ndone\n```\n\n For cleanup, I would avoid deleting the whole `Application Support/Codex` directory unless a full local reset is intended. A narrower recovery path is:\n  1. Quit Codex.\n  2. Stop any Codex Crashpad handler that is still writing to that Crashpad database.\n  3. Move only Crashpad report buckets such as `pending` ,`new` ,`completed` , and`attachments` to Trash.\n  4. Restart Codex and watch the pending count for a few minutes.\n Example Trash-only version: \n\n```\nosascript -e 'quit app \"Codex\"' 2>/dev/null || true\npkill -f \"browser_crashpad_handler.*codex.*/Crashpad\" 2>/dev/null || true\n\nfor crash in \\\n  \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad\" \\\n  \"$HOME/Library/Application Support/Codex/Crashpad\"; do\n  [ -d \"$crash\" ] || continue\n  mkdir -p \"$HOME/.Trash/Codex-Crashpad\"\n  for bucket in pending new completed attachments; do\n    [ -d \"$crash/$bucket\" ] || continue\n    find \"$crash/$bucket\" -mindepth 1 -maxdepth 1 -exec mv {} \"$HOME/.Trash/Codex-Crashpad/\" \\;\n  done\ndone\n```\n\n I would not attach `.dmp` files publicly; they may contain memory, local paths, prompts, or project state. Metadata like counts, size, oldest/newest timestamp, app version, and which path is still growing should be enough for most public triage.Disclosure: I maintain SafeDisk, a local-first storage cleanup/checklist project. I wrote up this specific Crashpad cleanup boundary here: [https://www.getsafedisks.com/codex-crashpad-pending-dumps-disk-space?utm_source=github&utm_medium=issue-comment&utm_campaign=codex-crashpad-disk-space&utm_term=codex-crashpad-pending-dumps&utm_content=openai-codex-25921](https://www.getsafedisks.com/codex-crashpad-pending-dumps-disk-space?utm_source=github&utm_medium=issue-comment&utm_campaign=codex-crashpad-disk-space&utm_term=codex-crashpad-pending-dumps&utm_content=openai-codex-25921)If someone wants a private second read before deleting app state, that page now has a single on-page Deep Cleanup request form after the read-only checks.\n13. My temporary macOS workaround: I’m clearing Codex’s pending Crashpad reports every 15 minutes to avoid the local buildup: \n\n```\n*/15 * * * * /usr/bin/find \"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\" -maxdepth 1 -type f -delete >/dev/null 2>&1 || true\n```\n\n Setup on macOS: \n `crontab -e` Add the cron line above, save, then verify: \n `crontab -l` This deletes only files directly inside: \n `~/Library/Application Support/com.openai.codex/web/Crashpad/pending`\n14. In case you also have that problem, there is yet another way this vibe coded app wastes disk space for me, see [#25667 (comment)](https://github.com/openai/codex/issues/25667#issuecomment-4758527276)\n15. I found a method which seems working well. clean up the folder as mentioned by [@Jolg42](https://github.com/Jolg42) , then set it read-only.\n I tested few hours. Now Codex cannot write any trash to this folder any more.`Crashpad/pending/` only\n\n``` bash\n#!/bin/bash\nset -euo pipefail\n\nTARGET=\"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\"\n\n[[ -d \"$TARGET\" ]] || { echo \"Folder does not exist: $TARGET\"; exit 0; }\n\necho \"Found $TARGET\"\n/usr/bin/find \"$TARGET\" -maxdepth 1 -type f -delete >/dev/null 2>&1 || true\nchmod -R a-w \"$TARGET\"\n\necho \"Cleaned and made read-only: $TARGET\"\n```\n\n `Crashpad/pending/` +`com.openai.codex.code_sign_clone`\n\n``` bash\n#!/bin/bash\nset -euo pipefail\n\nTARGET=\"$HOME/Library/Application Support/com.openai.codex/web/Crashpad/pending\"\n\n[[ -d \"$TARGET\" ]] || { echo \"Folder does not exist: $TARGET\"; exit 0; }\n\necho \"Found $TARGET\"\n/usr/bin/find \"$TARGET\" -maxdepth 1 -type f -delete >/dev/null 2>&1 || true\nchmod -R a-w \"$TARGET\"\n\necho \"Cleaned and made read-only: $TARGET\"\n\nROOT=\"/private/var/folders\"\nNAME=\"com.openai.codex.code_sign_clone\"\n\n/usr/bin/find \"$ROOT\" -type d -name \"$NAME\" 2>/dev/null | while IFS= read -r dir; do\n  echo \"Found $dir\"\n  chmod -R u+w \"$dir\" 2>/dev/null || true\n  /usr/bin/find \"$dir\" -mindepth 1 -delete 2>/dev/null || true\n  chmod -R a-w \"$dir\" 2>/dev/null || true\n  echo \"Cleaned and made read-only: $dir\"\ndone\n```\n\n16. One caution on the `chmod` / read-only workaround above: I would treat it as an emergency brake, not the default cleanup path.Making only `Crashpad/pending` read-only can stop an active dump loop, but it also prevents Codex from writing crash reports and may hide the signal maintainers need. I would keep the scope narrow and avoid making these broader paths permanently read-only:\n  - `~/Library/Application Support/Codex`\n  - `~/Library/Application Support/com.openai.codex`\n  - temporary `com.openai.codex.code_sign_clone` directories\n My safer order would be: \n  1. Quit Codex.\n  2. Move only Crashpad report buckets such as `pending` ,`new` ,`completed` , and`attachments` to Trash.\n  3. Restart Codex and watch whether `pending` starts growing again.\n  4. Use a read-only workaround only if it is actively refilling and you need to stop disk loss immediately.\n  5. Undo the read-only bit after the app/version issue is fixed.\n I updated the checklist here with the narrower warning and a small rollback snippet: \n [https://www.getsafedisks.com/codex-crashpad-pending-dumps-disk-space?utm_source=github&utm_medium=followup&utm_campaign=first-paid-cleanup&utm_term=codex-crashpad-pending-dumps-disk-space&utm_content=readonly-safety-followup](https://www.getsafedisks.com/codex-crashpad-pending-dumps-disk-space?utm_source=github&utm_medium=followup&utm_campaign=first-paid-cleanup&utm_term=codex-crashpad-pending-dumps-disk-space&utm_content=readonly-safety-followup)\n17. This is a serious problem, it is rapidly damaging hardware in computers where memory is not replaceable. It must be fixed asap. Reacted by Phuwanai Thummavet, kou kumotora, whatcall, WilsonPiper and Ahmed Khan\n18. Still reproducible in the current unified ChatGPT/Codex desktop app, with the newer Crashpad database path. ## Environment\n  - App: ChatGPT unified desktop app (`com.openai.codex` )\n  - App version: `26.715.70719`\n  - Build: `5650`\n  - macOS: `15.7.4` (`24G517` )\n  - Architecture: Apple Silicon / arm64\n  - Chromium framework: `150.0.7871.124`\n  - Crashpad database: `~/Library/Application Support/Codex/Crashpad`\n ## Impact`Crashpad/pending` reached approximately**33.4 GiB** . About**23.3 GiB** had been generated during the last three days. Deleting pending reports recovered the disk space, but new paired dumps/sidecars began appearing immediately.The main ChatGPT window can remain alive while `browser_crashpad_handler` crashes and respawns repeatedly. macOS retained at least 50 handler crash reports on 2026-07-20 and 43 on 2026-07-21; the real number of Crashpad pending dumps was much higher.## Exact handler failure\n\n```\n[FATAL:third_party/crashpad/crashpad/handler/mac/exception_handler_server.cc:76]\nCheck failed: kr == KERN_SUCCESS.\nmach_port_request_notification: (os/kern) invalid capability (20)\n```\n\n The macOS `.ips` reports show:\n\n```\nprocName: browser_crashpad_handler\nresponsibleProc: ChatGPT\nexception: EXC_BREAKPOINT / SIGTRAP\nprocPath: /Applications/ChatGPT.app/Contents/Frameworks/Codex Framework.framework/Versions/150.0.7871.124/Helpers/browser_crashpad_handler\n```\n\n The earliest retained report on this machine is from `2026-07-20 08:17:09 +0800` , before build 5650 was installed on this machine. Its handler UUID is`4C4C44FF-5555-3144-A167-DF0A5F297DB2` .I also downloaded the official build 5538 ( `26.715.31251` ) and compared its handler with build 5650. Both use Chromium`150.0.7871.124` , have the same Mach-O UUID, and have an identical SHA-256 hash for the unsigned code region. Therefore rolling back only to build 5538 is unlikely to avoid this failure.## Workarounds tested\n  - Launching with `--disable-breakpad` : did not prevent new Crashpad reports.\n  - Denying execution of the handler: stopped pending dumps but caused repeated macOS “ChatGPT quit unexpectedly” reports/popups, so this is not viable.\n  - Periodically deleting only `Crashpad/pending` : contains disk usage but does not fix the handler crash loop.\n No `.dmp` file is attached because it may contain private process memory. I can provide sanitized metadata or use a maintainer-provided private upload channel if needed.Could the desktop app either fix the handler startup/monitor-self failure, honor a supported crash-reporting opt-out, or enforce a strict retention/size limit on pending dumps?\n19. App: ChatGPT unified desktop app (\n20. Same issue here. The crazy thing is that a pruner does ship with Crashpad, but it's not implemented here. I've seen this behaviour in other Electron apps, where the handler is wired up but the pruner is not, and if telemetry is turned off or blocked, these logs will just pile up in pending.\n21. ### austinhutchison commentedon Aug 15, 2026 More actionsI’m seeing the same Crashpad pending-report growth on the current unified ChatGPT/Codex desktop app. Environment: \n  - App: ChatGPT desktop app with Codex\n  - App version: 26.803.61601\n  - Build: 6396\n  - macOS: 26.3.1 (25D771280a)\n  - Architecture: arm64\n The affected path is: ~/Library/Application Support/Codex/Crashpad/pending This directory grew to approximately 100 GB and roughly one million files. The files appear to be Crashpad pending reports, including `.dmp` files and sidecar metadata. The app remained usable while the directory grew.This appears to be the newer `Codex/Crashpad` path rather than the older:~/Library/Application Support/com.openai.codex/web/Crashpad/pending Deleting or moving the contents of `Crashpad/pending` recovers the disk space, but I have not yet established whether the directory regrows after cleanup.Could someone clarify whether the comment that “later macOS versions, at least 26.5,” fix this refers to a known macOS fix? If so, is macOS 26.5 expected to prevent the Crashpad handler failure, or is a Codex app update also required? I have not attached dump files because they may contain process memory, local paths, or other private data. I can provide sanitized counts, timestamps, or logs if that would help diagnose it.\n22. I am seeing what appears to be the same class of issue on macOS, although my Crashpad database path is different: ~/Library/Application Support/Codex/Crashpad/pending My Codex Desktop version is 26.825.51511. The live Codex helper processes use: --user-data-dir=/Users/xxx/Library/Application Support/Codex \n --database=/Users/xxx/Library/Application Support/Codex/CrashpadThe pending directory contains the same paired `.dmp` and`_sidecar.json` files described in this issue. Some sampled reports date back to June 2026, while others are from late August.The directory has become so large that even enumerating it is very slow: a bounded directory read did not finish within 8 seconds. A bounded scan managed to process only 118 entries (about 10 MB) before timing out, so this is not a normal small crash-report queue. It also appears to still be active rather than merely containing old leftovers: the `pending` directory modification time advanced from`2026-09-01 15:51:03 +0800` to`2026-09-01 15:51:34 +0800` during a 30-second read-only observation.Separately, when this happens I have also observed a process named `ReportClash` continuously consuming CPU. I do not yet know whether it is a Codex child/helper process or an unrelated process, so I am reporting it as a correlation only.\n23. Still present on the current buildת adding a data point with different characteristics from the browser_crashpad_handler SIGTRAP loop in [#36982](https://github.com/openai/codex/issues/36982) .Environment: ChatGPT.app 26.818.41509 (build 6962), bundle com.openai.codex, Codex Framework (Chromium) 151.0.7922.170, macOS 15.6.1. ~/Library/Application Support/Codex/Crashpad/pending is 174.6 GB, against 171 MB for everything else in Application Support/Codex combined — 99.9% of the directory. Uniform ~161 KB .dmp + _sidecar.json pairs ({\"capture_kind\":\"crash\"}), so ≈1.08M dumps and >2M files. The directory's link count has saturated at 65535 and the directory entry itself is 65 MB. Differences from [#36982](https://github.com/openai/codex/issues/36982) : No browser_crashpad_handler-*.ips in ~/Library/Logs/DiagnosticReports (30 reports present, none from the handler). Not in a tight loop: last write to pending/ was 16 hours ago and the app has run since without adding to it. Dump annotations reference Utility and extension process types. Crashpad/ was created 2026-04-05, so ~1.1 GB/day averaged rather than 5–17 GB/day.Whatever generates the dumps, nothing prunes them. The retention bug looks separable from any specific crash loop: even a low crash rate reaches 175 GB given five months. A size or age cap on pending/ would bound the damage independently of fixing the crashers.\n24. Still present on a much newer build, and I think I can supply the missing piece: **why the loop sustains itself.**### Environment\n  - Codex/ChatGPT.app `26.831.21537` , Codex Framework (Chromium)`152.0.7977.64`\n  - macOS Tahoe `26.3.1` (`25D771280a` ), Apple Silicon\n ### The dumps are of the crash handler itselfEvery `.dmp` in`Crashpad/pending` is a minidump of`browser_crashpad_handler` , not of the app or a renderer. Each one contains the same fatal error:\n\n```\n[FATAL:third_party/crashpad/crashpad/handler/mac/exception_handler_server.cc:76]\nCheck failed: kr == KERN_SUCCESS.\nmach_port_request_notification: (os/kern) invalid capability (20)\n```\n\n with annotations: \n\n```\nexecutable_path=/Applications/ChatGPT.app/Contents/Frameworks/Codex Framework.framework/Versions/152.0.7977.64/Helpers/browser_crashpad_handler\n--monitor-self-annotation=ptype=crashpad-handler\nptype=crashpad-handler\n```\n\n Because the handler is launched with `--monitor-self-annotation` , it monitors itself. So the sequence is:\n  1. handler calls `mach_port_request_notification` , gets`KERN_INVALID_CAPABILITY` (20)\n  2. hits the `CHECK` at`exception_handler_server.cc:76` and dies\n  3. the self-monitor writes a minidump **of the handler**\n  4. the handler is relaunched, fails the identical check, goto 2\n That's a loop with no base case, which is why the growth is perfectly linear rather than correlated with any app activity. ### Measured rateExactly **1 dump per second** ,`181,552` bytes each, plus a 24-byte`_sidecar.json` :\n  - ~3,600 dumps/hour ≈ **0.6 GB/hour**\n  - ≈ **15 GB/day** while the app is running\n In my case it reached **240,966 files / 21 GB** , and took a 228 GB disk down to**0 bytes free** — to the point that unrelated tooling could no longer create files.`completed/` was empty across all 240,966 reports, so nothing ever drains: uploads never succeed, and there is no cap or TTL on`pending/` .macOS independently logged 387 `.ips` reports for the same helper under`~/Library/Logs/DiagnosticReports` , oldest dated a week before I noticed anything. There is no user-visible signal at all until the disk fills.### The trigger is upstream, the amplification is notThe `mach_port_request_notification` /`KERN_INVALID_CAPABILITY` failure is not specific to this app — the identical error on Tahoe 26.3.1 is reported against other Chromium/Electron apps:\n  - [Claude Desktop crashes on macOS Tahoe 26.3.1 — Crashpad mach_port_request_notification failure anthropics/claude-code#37230](https://github.com/anthropics/claude-code/issues/37230)\n  - [Integrated terminal on macOS freezes permanently after running a program that segfaults — Crashpad fatal error kills PTY host microsoft/vscode#282433](https://github.com/microsoft/vscode/issues/282433)\n So the `CHECK` failure itself is a Crashpad/Tahoe Mach IPC problem. But the unbounded disk consumption is independent of it:*any* repeatable handler-startup failure produces the same runaway, because the self-monitor treats the handler's own fatal`CHECK` as a reportable crash.### Two independent mitigations\n  1. **Don't let the self-monitor report the handler's own startup `CHECK` failure** — or back off / rate-limit after N consecutive identical dumps. This alone breaks the loop regardless of the upstream cause.\n  2. **Bound `pending/`** with a size cap or TTL so a silently failing upload path cannot fill the disk.\n Either one would have turned this from \"disk at 0 bytes\" into a log line. ### WorkaroundEmptying `pending/` and then removing write permission from the directory stops the disk growth immediately (the handler keeps retrying, but can no longer write):\n\n```\nP=~/Library/\"Application Support\"/Codex/Crashpad/pending\nfind \"$P\" -type f -delete\nchmod 500 \"$P\"\n```\n\n Confirmed: growth stops, and nothing spills into `new/` ,`completed/` , or`attachments/` — the whole`Crashpad` directory stays at 20K. Reverse with`chmod 700` .Reacted by Glenn 'devalias' Grant\n25. Codex/ChatGPT.app \n26. Adding a measured data point from 2026-10-05 on macOS 26.0.1 (25A362), arm64. `~/Library/Application Support/Codex/Crashpad/pending` contained 1,212,428 files removed during cleanup: 110,728,502,872 logical bytes / 114,221,113,344 allocated bytes (114.22 GB). File mtimes ranged from 2026-07-27 through 2026-10-04. The dump-header timestamp histogram included 85,720 dumps on 2026-08-27 alone. This backlog was a major contributor to an almost-full 500 GB disk.I parsed ten retained minidump samples before deleting them. All ten named this main module: \n\n```\n/Applications/ChatGPT.app/Contents/Frameworks/Codex Framework.framework/Versions/151.0.7922.76/Helpers/browser_crashpad_handler\n```\n\n Each sampled dump was 182,624 bytes and identified ARM64. This corroborates reports here that the crash handler itself is being dumped, although the sample does not establish that every report had the same cause. The currently installed app is 26.930.21537 (build 12776), with framework 154.0.8037.93. The pending queue has stayed empty during today's cleanup while the app remains usable. I am reporting a historical backlog, not claiming that the current build is actively reproducing the loop or that the underlying bug is fixed. For local containment, I installed a LaunchAgent that runs every five minutes, expires reports older than 24 hours, and additionally caps each pending/completed queue at 64 report groups and 256 MiB. It groups matching UUID dump/sidecar files, gives fresh writes a two-minute grace period, and skips unknown files, symlinks, and files that change during inspection. This bounds retained diagnostic storage without making the database unwritable, but it does not repair a crashing helper. A built-in retention/size limit would prevent this degree of disk exhaustion independently of the handler failure. No raw minidumps are attached.\n27. Still happening on the current build, and at a much larger scale. Measured 2026-10-15 to 2026-10-17: **Environment**\n  - ChatGPT.app `26.1007.71326` (build`20398` ), Codex Framework (Chromium)`155.0.8059.40`\n  - macOS `15.7.4` , Apple Silicon (arm64)\n **What I found** Folder Size Files `~/Library/Application Support/com.openai.codex/web/Crashpad` 180 GB 2,251,765 `~/Library/Application Support/Codex/Crashpad` (almost all of it in`pending/` )143 GB 1,785,095 **Total****~323 GB****~4.04 million** That was about 35% of a 926 GB disk. Free space fell to 1.4 GB before I found it. Nothing in the app warned me. I only found it by running `du` over`~/Library` .The app kept working the whole time, as others have reported here. So the dumps build up silently in the background with no limit. **Cleanup** With the app still running, I renamed both `Crashpad` folders and deleted the renamed copies. Free space went from 1.4 GB to 50+ GB within minutes, and the delete is still running (millions of small files).**Asks**\n  1. Put a size or count cap on `Crashpad/pending` , with automatic pruning (see also[Add storage limits and automatic cleanup for Codex-generated data #51361](https://github.com/openai/codex/issues/51361) ).\n  2. Stop the handler crash loop that produces the dumps ([\\[app\\]\\[macOS\\]\\[26.727.51351\\] browser_crashpad_handler SIGTRAPs 2ms after launch, launchd respawns it, Crashpad/pending fills disk at ~17GB/day #36982](https://github.com/openai/codex/issues/36982) ).\n  3. Show a warning in the app when its own data goes over a few GB.\n28. ChatGPT.app \n\n## Metadata\n\n## Metadata\n\n### Assignees\n\n### Labels\n\n[appIssues related to the Codex desktop app](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22app%22)Issues related to the Codex desktop app\n\n[bugSomething isn't working](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22bug%22)Something isn't working\n\n[performance](https://github.com/openai/codex/issues?q=state%3Aopen%20label%3A%22performance%22)", "url": "https://wpnews.pro/news/codex-is-eating-my-mac-s-disk-space", "canonical_source": "https://github.com/openai/codex/issues/25921", "published_at": "2026-10-17 17:45:08+00:00", "updated_at": "2026-10-17 22:08:28.002000+00:00", "lang": "en", "topics": ["ai-tools", "ai-products", "artificial-intelligence"], "entities": ["OpenAI", "Codex Desktop", "Crashpad", "GitHub", "macOS"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/codex-is-eating-my-mac-s-disk-space", "markdown": "https://wpnews.pro/news/codex-is-eating-my-mac-s-disk-space.md", "text": "https://wpnews.pro/news/codex-is-eating-my-mac-s-disk-space.txt", "jsonld": "https://wpnews.pro/news/codex-is-eating-my-mac-s-disk-space.jsonld"}}