{
 "experiment": {
  "id": 2,
  "name": "kvmachines/claude-science"
 },
 "generated": "2026-10-10T13:30:28Z",
 "totals": {
  "pages": 29,
  "tested": 2,
  "passing": 2,
  "runs": 4
 },
 "pages": [
  {
   "n": 1,
   "slug": "overview",
   "id": "claude-science-01-overview",
   "page_url": "https://claude.com/docs/claude-science/overview",
   "state": "done",
   "title": "Claude Science \u00b7 Claude Science",
   "runs": [
    {
     "run_id": "33be6d00bdcd4f4d81b48f6d391796f0",
     "name": "overview-1",
     "status": "FINISHED",
     "started": "2026-10-10T11:53:58Z",
     "page": "overview",
     "url": "https://claude.com/docs/claude-science/overview",
     "version": "0.1.55",
     "passed": 0,
     "attempt": "1",
     "evidence": "Fetched overview.md (HTTP 200, 7375 B) after granting claude.com. Kernels: Python 3.11.16, R 4.5.3, bash 5.3.20 (page says 'Python, R, or shell'). Sandbox enforced: bind() EPERM, no DNS, egress proxy-only, private IPs hard-refused, /bin/ps denied. Artifacts versioned with provenance: host.artifacts() returned artifacts each with latest_version_id, lineage=true. Reviewer live: host.findings() -> count 0, capabilities.findings=true. macOS 27.2 >= required 13; 14 GiB free >= ~5 GB.",
     "fix": "version mismatch: /Applications/Claude Science.app Info.plist reads 0.1.55 (both CFBundleShortVersionString and CFBundleVersion), not the expected 0.1.60; the ~/.local/bin CLI the get-started runs used could not be stat'd from the sandbox (data dir is outside the grants), so if two installs coexist, confirm with 'claude-science --version' on the Mac. Also overview.md contains no version string at all, so 0.1.60 cannot be sourced from this page -- pin expected version from release notes. Correction to commit bffd59f: MLflow is NOT reachable from the sandbox over HTTP; this run was written to the sqlite store on disk instead.",
     "data_dvc_md5": "c889f54174f5979ca23b952c874f0747"
    },
    {
     "run_id": "46bcaac6ea124e27b25a66fea40226ed",
     "name": "overview-2",
     "status": "FINISHED",
     "started": "2026-10-10T12:26:52Z",
     "page": "overview",
     "url": "https://claude.com/docs/claude-science/overview",
     "version": "0.1.60",
     "passed": 1,
     "attempt": "2",
     "evidence": "Attempt 2, after the operator confirmed the version in the app. Settings > General > About reports Claude Science 0.1.60, which is the installation under test, so the version claim holds and all five of the page's claims pass: kernels Python 3.11.16 + R 4.5.3 + bash 5.3.20 ('Python, R, or shell'); sandbox enforced (bind EPERM, no DNS, proxy-only egress, private IPs hard-refused, /bin/ps denied); artifacts versioned with provenance (host.artifacts() each with latest_version_id, lineage=true); reviewer live (host.findings() count 0, capabilities.findings=true). Requirements also hold: macOS 27.2 >= 13, 14 GiB free >= ~5 GB.",
     "fix": "Attempt 1 (run 33be6d00, pass 0) read the version from /Applications/Claude Science.app/Contents/Info.plist, which is a SEPARATE STALE INSTALL at 0.1.55. Its own update setting was not checked -- the 'automatic updates are off' line in Settings sits under About for the running 0.1.60 install, not for that bundle. The running app is the ~/.local/bin CLI at 0.1.60. Left over: delete or update the stale 0.1.55 bundle so no future page test measures the wrong install. Harness: overview.md carries no version string, so pin the expected version from release notes, never from the page under test.",
     "data_dvc_md5": "c889f54174f5979ca23b952c874f0747"
    }
   ],
   "attempts": 2,
   "latest": {
    "run_id": "46bcaac6ea124e27b25a66fea40226ed",
    "name": "overview-2",
    "status": "FINISHED",
    "started": "2026-10-10T12:26:52Z",
    "page": "overview",
    "url": "https://claude.com/docs/claude-science/overview",
    "version": "0.1.60",
    "passed": 1,
    "attempt": "2",
    "evidence": "Attempt 2, after the operator confirmed the version in the app. Settings > General > About reports Claude Science 0.1.60, which is the installation under test, so the version claim holds and all five of the page's claims pass: kernels Python 3.11.16 + R 4.5.3 + bash 5.3.20 ('Python, R, or shell'); sandbox enforced (bind EPERM, no DNS, proxy-only egress, private IPs hard-refused, /bin/ps denied); artifacts versioned with provenance (host.artifacts() each with latest_version_id, lineage=true); reviewer live (host.findings() count 0, capabilities.findings=true). Requirements also hold: macOS 27.2 >= 13, 14 GiB free >= ~5 GB.",
    "fix": "Attempt 1 (run 33be6d00, pass 0) read the version from /Applications/Claude Science.app/Contents/Info.plist, which is a SEPARATE STALE INSTALL at 0.1.55. Its own update setting was not checked -- the 'automatic updates are off' line in Settings sits under About for the running 0.1.60 install, not for that bundle. The running app is the ~/.local/bin CLI at 0.1.60. Left over: delete or update the stale 0.1.55 bundle so no future page test measures the wrong install. Harness: overview.md carries no version string, so pin the expected version from release notes, never from the page under test.",
    "data_dvc_md5": "c889f54174f5979ca23b952c874f0747"
   }
  },
  {
   "n": 2,
   "slug": "get-started",
   "id": "claude-science-02-get-started",
   "page_url": "https://claude.com/docs/claude-science/get-started",
   "state": "done",
   "title": "Claude Science \u00b7 Get started",
   "runs": [
    {
     "run_id": "88a253340f2e4880b5a1a6f0ac52719a",
     "name": "get-started-1",
     "status": "FINISHED",
     "started": "2026-10-10T11:23:51Z",
     "page": "get-started",
     "url": "https://claude.com/docs/claude-science/get-started",
     "version": "0.1.60",
     "passed": 0,
     "attempt": "1",
     "evidence": "claude-science status: running pid 14123 port 8000; the web app opened on the login link and showed Account menu AKWLABS with akw-era projects",
     "fix": "sign the app out of the retired AKWLABS account and in with the KVMACHINES account (claudia profile), then run attempt 2",
     "data_dvc_md5": ""
    },
    {
     "run_id": "d20d535575ec426b9c0a5353dda97a65",
     "name": "get-started-2",
     "status": "FINISHED",
     "started": "2026-10-10T11:29:08Z",
     "page": "get-started",
     "url": "https://claude.com/docs/claude-science/get-started",
     "version": "0.1.60",
     "passed": 1,
     "attempt": "2",
     "evidence": "the web app's account menu reads KVMACHINES in the scientist profile tab; the Environments dialog listed python, r and compute-provider-modal as Ready",
     "fix": "",
     "data_dvc_md5": ""
    }
   ],
   "attempts": 2,
   "latest": {
    "run_id": "d20d535575ec426b9c0a5353dda97a65",
    "name": "get-started-2",
    "status": "FINISHED",
    "started": "2026-10-10T11:29:08Z",
    "page": "get-started",
    "url": "https://claude.com/docs/claude-science/get-started",
    "version": "0.1.60",
    "passed": 1,
    "attempt": "2",
    "evidence": "the web app's account menu reads KVMACHINES in the scientist profile tab; the Environments dialog listed python, r and compute-provider-modal as Ready",
    "fix": "",
    "data_dvc_md5": ""
   }
  },
  {
   "n": 3,
   "slug": "run-on-remote-linux-server",
   "id": "claude-science-03-run-on-remote-linux-server",
   "page_url": "https://claude.com/docs/claude-science/run-on-remote-linux-server",
   "state": "open",
   "title": "Claude Science \u00b7 Run on a remote server",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 4,
   "slug": "core-concepts",
   "id": "claude-science-04-core-concepts",
   "page_url": "https://claude.com/docs/claude-science/core-concepts",
   "state": "open",
   "title": "Claude Science \u00b7 Core concepts",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 5,
   "slug": "safeguards",
   "id": "claude-science-05-safeguards",
   "page_url": "https://claude.com/docs/claude-science/safeguards",
   "state": "open",
   "title": "Claude Science \u00b7 Safeguards",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 6,
   "slug": "multiple-computers",
   "id": "claude-science-06-multiple-computers",
   "page_url": "https://claude.com/docs/claude-science/multiple-computers",
   "state": "open",
   "title": "Claude Science \u00b7 Multiple computers",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 7,
   "slug": "artifacts",
   "id": "claude-science-07-artifacts",
   "page_url": "https://claude.com/docs/claude-science/artifacts",
   "state": "open",
   "title": "Claude Science \u00b7 Artifacts",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 8,
   "slug": "comments",
   "id": "claude-science-08-comments",
   "page_url": "https://claude.com/docs/claude-science/comments",
   "state": "open",
   "title": "Claude Science \u00b7 Comments",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 9,
   "slug": "the-reviewer",
   "id": "claude-science-09-the-reviewer",
   "page_url": "https://claude.com/docs/claude-science/the-reviewer",
   "state": "open",
   "title": "Claude Science \u00b7 The reviewer",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 10,
   "slug": "tools-and-environments",
   "id": "claude-science-10-tools-and-environments",
   "page_url": "https://claude.com/docs/claude-science/tools-and-environments",
   "state": "open",
   "title": "Claude Science \u00b7 Tools and environments",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 11,
   "slug": "troubleshooting",
   "id": "claude-science-11-troubleshooting",
   "page_url": "https://claude.com/docs/claude-science/troubleshooting",
   "state": "open",
   "title": "Claude Science \u00b7 Troubleshooting",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 12,
   "slug": "connectors-and-skills",
   "id": "claude-science-12-connectors-and-skills",
   "page_url": "https://claude.com/docs/claude-science/connectors-and-skills",
   "state": "open",
   "title": "Claude Science \u00b7 Connectors and skills",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 13,
   "slug": "remote-compute-clusters",
   "id": "claude-science-13-remote-compute-clusters",
   "page_url": "https://claude.com/docs/claude-science/remote-compute-clusters",
   "state": "open",
   "title": "Claude Science \u00b7 Remote compute clusters",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 14,
   "slug": "compute-providers",
   "id": "claude-science-14-compute-providers",
   "page_url": "https://claude.com/docs/claude-science/compute-providers",
   "state": "open",
   "title": "Claude Science \u00b7 Compute providers",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 15,
   "slug": "cloud-storage",
   "id": "claude-science-15-cloud-storage",
   "page_url": "https://claude.com/docs/claude-science/cloud-storage",
   "state": "open",
   "title": "Claude Science \u00b7 Cloud storage",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 16,
   "slug": "literature-access",
   "id": "claude-science-16-literature-access",
   "page_url": "https://claude.com/docs/claude-science/literature-access",
   "state": "open",
   "title": "Claude Science \u00b7 Literature access",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 17,
   "slug": "custom-connectors",
   "id": "claude-science-17-custom-connectors",
   "page_url": "https://claude.com/docs/claude-science/custom-connectors",
   "state": "open",
   "title": "Claude Science \u00b7 Custom connectors",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 18,
   "slug": "enable-claude-science",
   "id": "claude-science-18-enable-claude-science",
   "page_url": "https://claude.com/docs/claude-science/enable-claude-science",
   "state": "open",
   "title": "Claude Science \u00b7 Enable Claude Science",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 19,
   "slug": "how-claude-science-works-with-your-data",
   "id": "claude-science-19-how-claude-science-works-with-your-data",
   "page_url": "https://claude.com/docs/claude-science/how-claude-science-works-with-your-data",
   "state": "open",
   "title": "Claude Science \u00b7 How Claude Science works with your data",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 20,
   "slug": "admin-controls",
   "id": "claude-science-20-admin-controls",
   "page_url": "https://claude.com/docs/claude-science/admin-controls",
   "state": "open",
   "title": "Claude Science \u00b7 Admin controls",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 21,
   "slug": "manage-on-devices",
   "id": "claude-science-21-manage-on-devices",
   "page_url": "https://claude.com/docs/claude-science/manage-on-devices",
   "state": "open",
   "title": "Claude Science \u00b7 Manage on devices",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 22,
   "slug": "corporate-networks",
   "id": "claude-science-22-corporate-networks",
   "page_url": "https://claude.com/docs/claude-science/corporate-networks",
   "state": "open",
   "title": "Claude Science \u00b7 Corporate networks",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 23,
   "slug": "network-requirements",
   "id": "claude-science-23-network-requirements",
   "page_url": "https://claude.com/docs/claude-science/network-requirements",
   "state": "open",
   "title": "Claude Science \u00b7 Network requirements",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 24,
   "slug": "monitor-usage",
   "id": "claude-science-24-monitor-usage",
   "page_url": "https://claude.com/docs/claude-science/monitor-usage",
   "state": "open",
   "title": "Claude Science \u00b7 Monitor usage",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 25,
   "slug": "glossary",
   "id": "claude-science-25-glossary",
   "page_url": "https://claude.com/docs/claude-science/glossary",
   "state": "open",
   "title": "Claude Science \u00b7 Glossary",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 26,
   "slug": "command-line-settings",
   "id": "claude-science-26-command-line-settings",
   "page_url": "https://claude.com/docs/claude-science/command-line-settings",
   "state": "open",
   "title": "Claude Science \u00b7 Command line settings",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 27,
   "slug": "configuration-file-reference",
   "id": "claude-science-27-configuration-file-reference",
   "page_url": "https://claude.com/docs/claude-science/configuration-file-reference",
   "state": "open",
   "title": "Claude Science \u00b7 Configuration reference",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 28,
   "slug": "changelog",
   "id": "claude-science-28-changelog",
   "page_url": "https://claude.com/docs/claude-science/changelog",
   "state": "open",
   "title": "Claude Science \u00b7 Changelog",
   "runs": [],
   "attempts": 0,
   "latest": null
  },
  {
   "n": 29,
   "slug": "legal-and-compliance",
   "id": "claude-science-29-legal-and-compliance",
   "page_url": "https://claude.com/docs/claude-science/legal-and-compliance",
   "state": "open",
   "title": "Claude Science \u00b7 Legal and compliance",
   "runs": [],
   "attempts": 0,
   "latest": null
  }
 ],
 "orphan_runs": []
}