Skills

Implement features and bug fixes surgically, reusing existing code paths instead of adding parallel ones. Use when implementing a feature, fixing a bug, adding a new variant/harness/provider to an existing abstraction, or when the user says "surgical", "don't bloat", or "reuse what's there".

Set up and maintain a dedicated AI workspace, a folder separate from code where thinking artifacts compound across sessions. Use when the user wants to organize folders for AI work, asks where non-coding sessions should live, wants project insights extracted into their thinking folder, or complains that their agent keeps drowning in code-level context.

Speedrun a Foundation Sprint and derive a Founding Hypothesis from a coding session or project. Run this to turn a session/project into a product thesis and positioning — to define the customer, the problem, competitors and the 800-pound gorilla, classic and custom differentiators, the differentiation 2x2, the capability/insight/motivation advantage, values and guiding principles, product options scored through magic lenses, and a top bet. Use it whenever someone asks to "speedrun a foundation sprint", "derive a founding hypothesis", "find the product thesis", "figure out positioning", "who is the customer / what is the problem / who are the competitors", or "what makes this differentiated" — Claude participates in the sprint, brings its own ideas alongside the user's session-derived positions, and asks the human only what cannot be inferred.

Strategy for searching and reading the user's stored AI coding sessions efficiently. Use when the user asks broad questions about their history ("what have I been working on", "how did I solve X", "find sessions about Y") or when you need to triage many possible matches.

Shadow the user while they hand-edit code the agent just wrote, and distill their edits into style choices for future work. Use when the user says "keep an eye on what I'm doing", "learn my style from this", "watch my edits", or asks why their manual cleanup keeps repeating.

Script-first control for Ghostty on macOS. Open named terminals, run commands, drive TUIs, wait for rendered output, capture text or screenshots, manage splits/tabs, and clean up through composable gt-* scripts. Use when a task needs a real rendered terminal or terminal workspace.

Explains how decision traces are structured (moments, annotations, resolution) and how to summarize or reason about them. Use when the user asks about a specific trace, references a trace URL, asks to compare moments, or mentions decision history.

Drafts short, shareable tweets from a Skillsync session or trace that highlight tension, contradiction, or surprising moments. Use when the user asks for tweet ideas, social posts, share-worthy moments, "a fun tweet," "a punchy take," or any framing that names tension / contradiction / surprise / "didn't see that coming" from a specific session or trace.