Skip to content

Distinguish an unreadable source from a deleted one

Use explicit availability states rather than guessing why a source can no longer be inspected.

Guides

Distinguish an unreadable source from a deleted one — what should the operator record?

A failed source read can reflect a temporary problem, an access change or a removed object. Those conditions call for different operational responses. Store what the authorized response actually says, and resist the temptation to convert uncertainty into a definitive deletion claim that later tables inherit without review.

Practical operations check
Use explicit availability states rather than guessing why a source can no longer be inspected.
Save the returned error category and observation time in an access-controlled ledger, separating it from the last known descriptive source fields. Retain the time and actual returned category so a future authorized check can distinguish a new failure from a stale annotation copied between table revisions.
Explain whether availability recovered or remains unresolved, without claiming that a transient error proved removal or that restored access guarantees future availability. Avoid reporting inaccessible as false or unavailable as zero. Those shortcuts alter the meaning of evidence and can spread into summaries long after the transient issue ends.
01 / Guides

Capture the observed condition

Save the returned error category and observation time in an access-controlled ledger, separating it from the last known descriptive source fields. Retain the time and actual returned category so a future authorized check can distinguish a new failure from a stale annotation copied between table revisions.

02 / Guides

Choose a supported state

Use unavailable-unknown when no reliable reason is returned; reserve deleted or withdrawn for evidence that actually supports that narrower claim. A redirect or missing page can require review without establishing deletion; use the narrowest supported state until current evidence justifies a more definitive classification.

03 / Guides

Gate table use

Prevent rows needing current source validation from passing the release check while allowing clearly labeled historical metadata only where permitted. If a release requires current verifiability, separate blocked rows from admissible rows and preserve the reason rather than silently dropping them from the denominator.

04 / Guides

Review recovery evidence

If the source becomes readable through an authorized path, compare identifiers before restoring its current visibility state; a similar replacement URL may name another object. Restored readability should produce a new observation reference; it does not prove that the content stayed unchanged throughout the interval when the source was unreadable.

05 / Guides

Close the exception accurately

Explain whether availability recovered or remains unresolved, without claiming that a transient error proved removal or that restored access guarantees future availability. Avoid reporting inaccessible as false or unavailable as zero. Those shortcuts alter the meaning of evidence and can spread into summaries long after the transient issue ends.

Keep the boundaries clear.

  • Use explicit availability states rather than guessing why a source can no longer be inspected.
  • Operational record: Capture the observed condition. Context badges do not certify access.

Start with a read-only check: ask your agent to report actual source coverage before collecting records or exporting an audience.

Primary references

Documentation reviewed 2026-11-15. Product capabilities, client availability and plans can change. Third-party marks identify their products, not an affiliation.

Know what social media shows about you.
Start with a question.

Get started
montage