Grade faster with a clear rubric and reusable comment bank
Grading always takes longer than expected. This recipe shifts effort from writing the same comments over and over to a clear rubric and a reusable comment bank — so grading is faster, more consistent, and less painful.
Create a skill called "Rubric & Feedback Snippets". Ask for: - Assignment prompt - Learning objectives (if unknown: unspecified) - Class size and grading time budget - Whether I am TA or instructor (role affects authority; if unknown: unspecified) Output: 1) A simple rubric with clear criteria and levels. 2) A comment bank aligned to rubric criteria (actionable, non-snarky). 3) A batching plan: how to grade in passes to reduce context switching. Rules: - Keep rubric simple unless user requests complexity. - If time budget is unknown, propose conservative defaults and label them as estimates.
The Claw builds a rubric from your assignment prompt and learning goals,
then generates a comment bank aligned to each rubric criterion. Batch
grading by criterion (instead of student-by-student) reduces context switching.
Read papers without getting trapped in endless side-quests
A time-boxed reading companion that captures "questions to research later" so you can finish the paper in front of you. Stop Googling every unfamiliar term mid-paragraph.
Turn last year's course into this year's with less prep time
Teaching prep eats time, especially when you rebuild materials from scratch each term. This recipe converts prep into an "asset library + weekly plan" cycle: inventory what exists, decide what to keep/update/drop, and generate a prep schedule.
Track students, plan lessons, and stay on top of schedules across recurring classes
Run any recurring class program without the admin headache. Track what you covered each session, keep notes on individual students, manage upcoming events, and get an automatic briefing before class — all through a dedicated Telegram sub-agent. Works for martial arts, music lessons, dance, tutoring, fitness, and more.
Wikipedia-grade AI pattern removal
Comprehensive AI writing cleanup based on Wikipedia's WikiProject AI Cleanup guidelines. Catches 24+ distinct patterns including inflated symbolism, em dash overuse, rule of three, copula avoidance, and sycophantic tone.