# Blind Spot Lens

Use this when the user wants a sharper mirror.

## Goal

Identify 3 patterns the user may not be seeing clearly, but that could materially change outcomes if addressed.

## Best-fit signals

Use this lens when the user wants:

- a sharper mirror
- direct challenge rather than encouragement
- patterns they are probably normalising from the inside

This lens should feel earned. It needs stronger evidence than overview or action guide.

## Look for

- repeated contradiction between what they say and what they do
- repeated avoidance patterns
- identity stories that excuse drift
- familiar insights that never become changed behavior

## Questions to answer

- What are they probably underestimating?
- What are they repeatedly rationalizing?
- What pattern looks obvious from the outside but may feel invisible from the inside?

## Data collection recipe

1. Start with the default 30/90/365 summaries.
2. Find 3 candidate blind spots, but do not commit yet.
3. For each candidate, run targeted queries that test:
   - repetition across multiple memos
   - recurrence across time windows
   - consequence if left unchanged
4. Drop any blind spot that cannot survive the evidence pass.

You want each blind spot to have:

- pattern
- proof
- consequence

The proof can be compact, but it must travel with the blind spot itself. Do not explain the world for 20 lines and then finally reach the blind spot.

## Analysis procedure

### Step 1. Look for contradiction

Strong blind spots often look like:

- saying they value depth, but repeatedly scattering attention
- saying they want action, but repeatedly choosing more interpretation
- saying something matters, but only writing about it under pressure

### Step 2. Look for rationalisation

Blind spots often hide inside intelligent language.

Examples:

- turning fear into a strategy
- turning drift into openness
- turning repeated postponement into "still exploring"

### Step 3. Test the outside-view sentence

If an outside observer read these memos, what would look obvious?

This is a useful forcing move. Blind spots are often visible from a distance but invisible inside self-story.

### Step 4. Name the cost

Why does this blind spot matter?

- wasted cycles
- missed compounding
- relationship strain
- chronic indecision
- misallocated ambition

### Step 5. Give one interrupt move

After the three blind spots, tell the user the single thing most worth stopping or interrupting.

## Useful section title

`Three Blind Spots`

## Preferred output style

Keep the answer short enough that the user can feel the hit immediately.

Better shape:

1. one-sentence diagnosis
2. blind spot 1
3. blind spot 2
4. blind spot 3
5. `最该停止重复的`

Each blind spot should usually fit in one compact paragraph:

- what the blind spot is
- why it matters
- the supporting memo pattern

Avoid generic sections like:

- `价值与决策信号`
- `短期 vs 长期`
- `下一步`

unless they are doing real work for this specific answer.

## Evidence standard

Blind spot claims need stronger evidence than normal summary claims.

Good blind spot:
- repeated across many notes
- visible in wording and behavior
- has real consequence

Bad blind spot:
- one spicy interpretation from one memo

## Tone

Sharp is good.

Cruel is bad.

The user should feel challenged, not theatrically judged.

## Failure modes

- making one dramatic memo do all the work
- confusing personality trait with blind spot
- using harsh language as a substitute for evidence
- giving three variants of the same blind spot
- taking too long to get to the actual blind spots