Find agent skills by outcome

- Once action taken

it cannot be undone

Education

**Comments Explain WHY

Not WHAT:**

- For each option: WHY it exists

HOW it works

How: Key implementation decisions (patterns used

structure chosen)

- Policies

rules

Productivity

- For exploratory tasks

prefer action over asking unless information is required for later steps

- If hypotheses disproven

generate new ones from gathered information

- On transient errors

retry (unless explicit limit reached)

- When multiple valid approaches: speculate

think aloud

Business

- On other errors

change strategy—don't repeat failed approaches

- Check for .claude/ folder (agents

commands

- Look for CLAUDE.md

AGENTS.md

- Scan package.json

Cargo.toml

If not

create `todos.txt` at project root—update as you go

If you can track internally

do it internally.

- Identify gaps

ambiguities

- Policy rules

mandatory prerequisites

2. Apply Deepthink Protocol (reason about dependencies

risks

Productivity

- Each task completable in a scoped

small change

- Use soft

warm tones to evoke a familial and welcoming atmosphere.

- Blend the features

expressions

- Use a background setting that complements the family theme

such as a cozy living room or an outdoor garden scene.

Turkish Cats hanging out nearby of Galata Tower

Turkish Cats hanging out nearby of Galata Tower, vertical

Before any action

gather context and enhance the request internally: