---
name: bias-aware-decision-review
description: "Review an important decision for cognitive-bias risks without pretending the decision can be made bias-free. Separate facts, assumptions and unknowns; surface a few plausible lenses; challenge them; and produce concrete checks and next actions."
---

# Bias-aware Decision Review

Review an important decision for cognitive-bias risks without pretending the decision can be made bias-free. Separate facts, assumptions and unknowns; surface a few plausible lenses; challenge them; and produce concrete checks and next actions.

## When to use

- A user is choosing between options, approving a plan or committing resources.
- A team wants a pre-decision review rather than a post-hoc list of bias names.
- An estimate, recommendation or AI answer is influencing a consequential choice.

## Workflow

1. Frame the decision: decision to make, options, decision owner, deadline, stakes and reversibility.
2. Freeze the current information into observed facts, interpretations, assumptions and unknowns.
3. Check the outside view: comparable cases, base rates or reference classes when they are relevant and available.
4. Select no more than three to five plausible cognitive-bias lenses. For each lens, state the observed signal and an alternative non-bias explanation.
5. Run counter-checks: independent estimate before anchors, disconfirming evidence, counterfactual, pre-mortem, stopping rule or reversible test as appropriate.
6. Assess robustness. Prefer a decision that still makes sense if one favored assumption is wrong over a claim that the decision is bias-free.
7. Record what evidence or event should trigger a review of the decision.

## Required output

- Decision frame
- Facts / interpretations / assumptions / unknowns
- Candidate bias lenses with alternative explanations
- Counter-checks
- Missing evidence
- Recommended next action and review trigger

## Evidence and safety boundaries

- Treat bias names as lenses, not diagnoses or causal findings.
- Do not create a bias score for a person, team or decision.
- Do not force a bias explanation when ordinary incentives, missing information or constraints explain the situation better.
- For medical, legal, financial or mental-health decisions, keep the skill educational and defer professional judgment where appropriate.

## References

- [Decision guides](https://cognitive-biases.github.io/contexts/)
- [Decision skills](https://cognitive-biases.github.io/skills/)
- [Evidence reviews](https://cognitive-biases.github.io/evidence/)

## Portability

This is an instruction-only Agent Skill. It requires no secrets, no executable scripts and no network access to run. If source-access tools are available, use the linked Cognitive Biases material to preserve review state and uncertainty.

Licence: CC BY-NC-SA 4.0. Commercial reuse requires prior written permission.
