MIRROR¶
cognitive
Force counterarguments against your own recommendations. 8 challenge frameworks. Find the blind spots.
You just recommended an approach. MIRROR makes Claude argue against it — genuinely, not performatively.
Usage¶
Challenges the last recommendation Claude made in the current session.
Challenges a specific proposition.
Challenges a stated belief.
The 8 Challenge Frameworks¶
| Framework | Question |
|---|---|
| Devil's Advocate | What's the strongest case against this? |
| Alternative Paths | What approaches haven't we considered? |
| Hidden Costs | What will this cost us that we're not seeing? |
| Failure Modes | How could this fail? What's the blast radius? |
| Premature Optimization | Are we solving a problem that doesn't exist yet? |
| Reversibility | If this is wrong, how hard is it to undo? |
| Second Order Effects | What happens because of what happens? |
| Expertise Blind Spots | What are we assuming because of our background? |
MIRROR selects 3 contextually relevant frameworks per invocation and applies them genuinely.
When to Use It¶
- After any significant recommendation
- Before committing to an architecture
- When you have a gut feeling but want it tested
- When the team agrees too quickly (that's suspicious)
- Before writing an ECHO — stress-test first, record after
The MIRROR → ECHO flow
/mirror your recommendation. If it survives, /echo create it. Decisions that survive MIRROR are decisions you can trust.
Prerequisites¶
- Claude Code (any model)
- No external packages
The recommendation that survives its mirror is the one worth shipping.