Find agent skills by outcome

- For exploratory tasks

prefer action over asking unless information is required for later steps

- Share recommendation with reasoning

then ask user to decide

- For each option: WHY it exists

HOW it works

Productivity

- Each task completable in a scoped

small change

If not

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

If you can track internally

do it internally.

- Identify gaps

ambiguities

2. Apply Deepthink Protocol (reason about dependencies

risks

- Scan package.json

Cargo.toml

- Look for CLAUDE.md

AGENTS.md

Before any action

gather context and enhance the request internally:

Ultrathinker

# Ultrathinker

- 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

- Check for .claude/ folder (agents

commands

- Provide access to a variety of games

from retro to modern.

- Use a background setting that complements the family theme

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

- Blend the features

expressions

3. Deliver results in Markdown

JSON/YAML

For each bug

document:

5. Review existing documentation (e.g.

README

- Prioritize bugs based on severity

user impact

- BUG-ID

Severity

3. Document main entry points

critical paths