Kept AI
Back

Analytics & Monitoring

Check whether AI answers reflect a launch

Review recent answer samples for launch facts and clearly report timing limits.

Connect your Kept MCP
Kept AIAI assistantYour workspace
Open Kept AI

Command

Use list_projects({}) to confirm my supplied project host is visible. Use my chosen days value of 7, 30, or 90 for every time-windowed call. Read get_overview({host, days}) and search_chats({host, query: myLaunchTerm, days, limit: 20}). Use list_prompts({host, limit: 100}) and get_prompt_runs({host, prompt_id, days}) to verify launch claims against my supplied announcement facts. Inspect returned timestamps for post-launch samples, including a 48-hour subset only when supported. Compare with my saved pre-launch export where compatible. Report observed adoption, outdated statements, and insufficient evidence separately. Do not request custom overview windows or call the launch successful from visibility alone. Return evidence links, coverage limits, and explicit unknowns; never infer unsupported scores.

What you need

  • A visible host, supported window, launch timestamp, search term, and verified announcement facts.
  • A saved pre-launch export for comparisons; a 48-hour check depends on returned sample timestamps.
Read the setup guide

Check launch facts in observed answers

Search for the supplied launch term and compare sampled claims with the announcement facts. Use returned timestamps to identify relevant observations, explaining when the available samples cannot support a narrow post-launch view or a confident verdict about how broadly the update has spread.

Report adoption and uncertainty separately

Summarize which facts appear correctly, which statements remain outdated, and where evidence is absent. Keep visibility observations separate from commercial launch outcomes, and use compatible saved baseline material when available to explain changes without inventing an exact historical comparison window.