All runbooks
evidenceuai-explorerredacted view

MacBook-Air Sanitized Export Evidence

Synced UAI Explorer evidence documentation from docs/evidence/sanitized-export-macbook-air-2026-06-20.md.

Updated 2026-06-21·Freshness: Reference·Sensitive tokens or credentials were hidden before display.

MacBook-Air Sanitized Export Evidence

Date: 2026-06-20
Timezone: America/St_Johns

Scope

Created a metadata-only survey and sanitized export for MacBook-Air. The survey used SSH BatchMode as non-root scott against private IP 10.0.0.83; the host reported scotts-MacBook-Air.local and Darwin. The survey used explicit Desktop and Documents roots only.

This did not mount MacBook source folders into Explorer, copy raw source files into the public container, quote file contents, mutate production CT 121, mutate sandbox CT 126, or change public routes. Temporary survey worker files and output files under remote /tmp were removed after metadata artifacts were copied back.

Survey Command Shape

scp tools\survey_device_documents.py tools\sanitize_survey_export.py `
  scott@10.0.0.83:/tmp/uai-explorer-survey-worker-20260620/

ssh scott@10.0.0.83 "/bin/sh /tmp/uai-run-macbook-survey-20260620.sh"

The temporary remote runner executed:

python3 survey_device_documents.py \
  --device MacBook-Air \
  --host "$(hostname)" \
  --target "scott@10.0.0.83" \
  --access-outcome "SSH BatchMode connected to MacBook-Air as non-root scott; bounded metadata-only documentation survey completed against explicit roots." \
  --root "Desktop=$HOME/Desktop" \
  --root "Documents=$HOME/Documents" \
  --out /tmp/uai-macbook-air-survey-20260620 \
  --summary /tmp/macbook-air-2026-06-20-summary.md \
  --date 2026-06-20 \
  --max-depth 5 \
  --max-entries 6000 \
  --max-entries-per-root 3000

Sanitized Export Command

python tools\sanitize_survey_export.py `
  --manifest raw-survey-zone\macbook-air\2026-06-20\manifest.json `
  --inventory raw-survey-zone\macbook-air\2026-06-20\inventory.csv `
  --out sanitized-exports\macbook-air\2026-06-20

Outputs

  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].csv
  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].md
  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].json
  • [REDACTED:high-entropy].md

Verification

  • Survey candidate docs seen: 899.
  • Recorded metadata rows: 890.
  • Secret-adjacent aliases suppressed before output: 9.
  • Sanitized exported review items: 890.
  • Topic groups: 9.
  • Roots represented: Desktop and Documents.
  • Inventory truncated: no.
  • Notable topic counts: UAI 316, desktop documentation candidates 217, verification and reports 156, Cadence 71, Cadence/UAI Orb 39, AI delegation notes 30, infrastructure operations 30, UI/UX standards 17, and Cortex/Skipper operations 14.
  • raw_content_included=false.
  • raw_file_aliases_included=false.
  • raw_source_hashes_included=false.
  • public_publishable=false.
  • No generated review item or catalog candidate is marked publishable.
  • Scans over the sanitized export found no raw file aliases, raw source hashes, or secret-shaped values.
  • Scans over the raw metadata manifest, inventory, and public summary found no secret-shaped values.
  • Remote /tmp worker, runner, summary, and output paths were absent after cleanup.

The sanitized export is safe for outside-agent classification prompts, but it is not approved public catalog content. Raw lookup still requires private operator access to the original survey inventory.