Teach and explain existing codebases at project, module, and deep source-reading scopes. Use when the user asks to learn a repository, module, subsystem, feature area, or source implementation; trace call paths; understand AI/ML code with tensor shapes and formulas; continue a saved code-learning plan; or plan a…
Teach and explain existing codebases at project, module, and deep source-reading scopes, with source-grounded call paths, tensor shapes, formulas, and persistent learning plans when useful.
Turn an empirical research question or tentative idea into an evidence-backed paper through literature discovery, novelty and feasibility checks, adversarial clarification, persistent project state, staged experiments, claim-evidence mapping, drafting, and review. Use when the user asks to discover or validate a…
Teach, explain, and verify a specific academic paper or technical report through source-grounded structural mapping, prerequisite repair, equation and figure walkthroughs, small reproducible checks, misconception diagnosis, active recall, and resumable reading notes. Use when the user asks to summarize, read, learn…
Use only when the user explicitly invokes $ponytail-lite or explicitly says to use the ponytail-lite skill. Do not auto-select this skill from ordinary coding, review, refactoring, debugging, simplification, YAGNI, or over-engineering requests. Once explicitly invoked, apply lazy senior developer mode for the current…
★not rated 2 2mo agoA107 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: