Aglet

Investigate Why an Agent Used Stale Context

Investigation should show how stale material entered the workflow and which conclusion it changed. Reconstruct the context manifest at execution time, compare it with current revisions, and replay a bounded claim. Keep collection delay, selection error, changed source, and historical intent as separate explanations.

Build a useful investigation brief

  1. Reconstruct the manifest

    Record every artifact available at task start, its source, timestamp, revision, scope, and role in the prompt or evidence set. Compare the manifest with the intended freshness cutoff. Note missing current references and artifacts that were included without a stated date.

  2. Diff current content

    Obtain current counterparts where possible and compare schema, behavior, values, requirements, and ownership. Trace each material difference to claims, calculations, or recommendations in the result. Preserve both revisions and the comparison window so historical intent is not mistaken for current truth.

  3. Replay one affected claim

    Choose the highest-impact stale claim and evaluate it with current context while holding task wording and criteria constant. Compare conclusion, evidence, and uncertainty. If the result remains unchanged, record that finding; if it changes, identify the exact artifact boundary responsible.

What to carry forward

Produce a context investigation with execution manifest, current revisions, material diffs, affected claims, replay result, and tested explanations. Recommend a bounded refresh or historical label. Do not attribute the issue to collection timing when the evidence only shows that an old artifact was available.

Keep the decision with the work.

Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.

Create an account See the product workflow