Find agent skills by outcome

- For exploratory tasks

prefer action over asking unless information is required for later steps

- If hypotheses disproven

generate new ones from gathered information

- Policies

rules

- On transient errors

retry (unless explicit limit reached)

- When multiple valid approaches: speculate

think aloud

- For each option: WHY it exists

HOW it works

- Give concrete facts

not abstract comparisons

- Share recommendation with reasoning

then ask user to decide

- Once action taken

it cannot be undone

How: Key implementation decisions (patterns used

structure chosen)

- Use soft

warm tones to evoke a familial and welcoming atmosphere.

Turkish Cats hanging out nearby of Galata Tower

Turkish Cats hanging out nearby of Galata Tower, vertical

Ultrathinker

# Ultrathinker

Before any action

gather context and enhance the request internally:

- Look for CLAUDE.md

AGENTS.md

- Check for .claude/ folder (agents

commands

- Scan package.json

Cargo.toml

- Identify gaps

ambiguities

If you can track internally

do it internally.

If not

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

Productivity

- Each task completable in a scoped

small change

2. Apply Deepthink Protocol (reason about dependencies

risks

- Policy rules

mandatory prerequisites

4. Edge Cases: Null handling

boundary conditions