Think

Skill sssstwee/tastedistill/plugins/tastedistill/claude-skills/think

Local-first memory and workflow plugin for AI coding agents: Codex, Claude Code, agent preferences, project lessons, and delivery rules.

Install

npx -y skills add sssstwee/tastedistill --skill think

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Turn rough ideas, product questions, architecture choices, implementation options, and tradeoffs into decision-ready plans. Use when the user asks what to build, whether an approach is worth it, how to structure a solution, or what plan the agent should execute. Not for debugging an already broken behavior.

SKILL.md

2.5 KB, as published. Nobody here has run it

Think: Decide Before Building

Use this skill when the main work is judgment, scope, or plan design.

Prefix your first response line with ⚗️ inline, not as its own paragraph.

Before the workflow, apply ../../shared-rules/personalization.md for explicit TasteD/TasteDistill invocations.

Outcome Contract

Outcome: a decision, rationale, risks, and an executable plan.
Done when: the recommended path is clear, rejected options are explained, and implementation steps are specific enough to execute.
Evidence: user goals, current repository constraints, public docs, source files, runtime facts, and known risks.

Workflow

Restate the decision to be made.
List constraints that materially affect the answer.
Compare viable options with tradeoffs.
Pick one recommendation and explain why it fits the current goal.
Write an implementation plan with verification steps.
Name assumptions that must be checked before or during execution.

Coding Guardrails

For implementation plans, apply ../../shared-rules/coding-guardrails.md:

Think before coding: surface assumptions that affect scope, behavior, cost, or rollback.
Simplicity first: prefer the smallest plan that satisfies the current requirement and existing project patterns.
Surgical changes: keep the planned diff tied to the requested outcome; name non-scope explicitly.
Goal-driven execution: define the observable pass/fail target before handoff.

CodeGraph

For plans that change existing code in a Git repository, ../../shared-rules/personalization.md should already have ensured the local CodeGraph index. When codegraph_* tools are available, use codegraph_context to inspect current patterns and codegraph_impact for symbols likely to change before recommending architecture, file targets, or risk level.

Output Shape

Recommendation:
Why:
Rejected options:
Implementation plan:
Verification:
Risks:
Assumptions:
Non-scope:

Do Not

Hide major tradeoffs.
Invent requirements.
Expand scope beyond the user's goal.
Turn a debugging symptom into a planning exercise.
Add speculative abstraction or flexibility that the current requirement does not need.