AI agent for director of strategies
Strategic Assumption Watch Agent
Leadership always knows which strategic assumptions still hold and what changes when one breaks.
What it does
Strategic plans rest on assumptions, such as market growth, a competitor's pricing or interest rates. When those change, plans often stay the same. This agent runs monthly or when new source data arrives. It ranks assumptions by how much their evidence has weakened, reads approved sources for the weakest one, and records the finding in a dated evidence ledger. If new evidence conflicts with earlier findings, it records the conflict and asks the owner instead of picking a side. When a threshold is crossed with supported evidence, it recalculates the affected scenario and opens a decision task for the owner. The strategy head decides on plan changes. Edge case: a single unverified news report never crosses a threshold on its own.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Monthly review or new source data
- Rank assumptions by how much evidence has weakened
- Read approved sources for the weakest assumption
- Is the new evidence consistent with earlier findings?If not: record the conflict and ask the owner. Back to step 3.
- Is the threshold crossed?If not: update the ledger and move to the next assumption. Back to step 2.
- Recalculate the affected scenario
- Open a decision task for the owner
- Strategy head decides on plan changesThe agent waits here for your OK.
- Current assumption register and scenario update
How it decides
It picks the next source for the weakest assumption, compares new evidence with dates and prior findings, recalculates only when a supported threshold is crossed, and asks the owner when evidence conflicts.
- Which assumption next: the weakest or most decision-critical.
- Recalculate: only when a threshold is crossed with supported evidence.
- Conflict handling: show both sources, never choose silently.
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Approved sources per assumption
- Threshold per assumption that triggers a recalculation (default set by the owner)
- Review schedule (default monthly)
- Who owns each assumption and receives questions
- Scenario model used for recalculation (default the plan's finance model)
What keeps you in control
It always asks you first
- Changing strategy
- Budgets
- External commitments
Hard limits
- Approved sources only.
- No plan changes itself.
It stops when
- Done: all assumptions reviewed.
- Needs a human: conflicting evidence.
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide