Skip to content

Map dependencies between research issues without assuming causation

Identify when resolving one documented obstacle may affect another, while keeping observed sequences separate from untested causal relationships.

Guides

Are two reported problems separate tasks or a possible chain?

Public accounts of photo library migration may mention failed imports followed by missing albums. Those issues might be linked, but the same words can also describe separate problems. A dependency map helps a team investigate the possible sequence without assuming that fixing the first issue will necessarily remove the second in every workflow or account.

Example research prompt
Map supplied photo migration issues into documented obstacles and explicitly labeled possible dependencies. Seek separating cases and do not infer causation from co-occurrence alone.
01 / Guides

Extract explicit ordering

Record whether the author describes one event before another and whether the relationship is stated or inferred. Include the relevant task sequence and conditions. Exclude a causal arrow based only on two issues appearing in the same paragraph; co-occurrence is not enough to establish either ordering or a mechanism connecting the reported outcomes.

02 / Guides

Classify the proposed link

Use labels such as stated dependency, observed sequence and analyst hypothesis. Keep independent issues unconnected unless evidence supports a relationship. If the author speculates that an import failure caused missing albums, preserve that attribution rather than treating it as a verified technical diagnosis or a confirmed dependency in the final issue graph.

03 / Guides

Seek separating cases

Look for missing albums after a reportedly successful import and for failed imports without the album issue. These cases can test whether the proposed chain is necessary or sufficient within the reviewed material. Keep missing details unresolved, and do not discard separating cases merely because they complicate an attractive single-root-cause explanation for the team.

04 / Guides

Deliver an investigation graph

Provide nodes for documented obstacles and labeled links for their evidential status. State which issue should be investigated first because of a plausible dependency, not because a causal relationship has been proven. If the evidence cannot support links, deliver a grouped issue list instead of drawing precise arrows that imply a stronger model than the public accounts establish.

Keep the boundaries clear.

  • A useful investigation order does not require a proven causal chain.
  • Research-context badges do not promise source access, complete history or live verification.

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 2026-11-13. 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