Kept AI
Back

Strategy & Research

Find competitor advantages by prompt

Inspect tracked prompts where a named competitor has stronger returned evidence.

Connect your Kept MCP
Kept AIAI assistantYour workspace
Open Kept AI

Command

Use my supplied host, competitor name, and days of 7, 30, or 90. Call list_projects and verify the host, then list_prompts({host,status:"active",limit:100}), following nextCursor with cursor until complete. Read get_gap_analysis({host,days}). Select relevant prompt IDs and inspect get_prompt_runs({host,prompt_id,days}); use search_chats({host,query,days,limit:20}) with the competitor name for context. Build a table of prompt, observed competitor advantage, brand evidence, citations, and sample coverage. Rank only using returned metrics or clearly stated qualitative criteria. Mark unavailable scores unknown, avoid extrapolating to untracked queries, and propose three prompt-specific content investigations for human review.

What you need

  • Supply a visible host, a named competitor, and a supported window.
  • Allow prompt pagination; conclusions cover tracked prompts and returned samples only.
Read the setup guide

Inspect the prompts behind competitor gaps

Connect competitor gap summaries to specific tracked prompts and recent sampled answers. The table preserves citations and coverage limits, so a competitor mention is not automatically treated as a win and a missing brand mention is not generalized beyond the observed samples.

Prioritize evidence-supported investigations

Choose a few prompts where the available evidence suggests a useful content investigation. Explain the ranking criteria and the question each investigation should resolve, giving the team an actionable review queue without inventing numeric opportunity scores or guaranteed visibility gains.