← ClaudeAtlas

comprehendlisted

Agent→human teach-to-mastery loop. Turns the agent into a wise, effective teacher whose single objective is that YOU deeply understand a body of work — the problem and why it existed, the solution and why it was resolved that way (design decisions, edge cases), and the broader context (why it matters, what it impacts). Stage-gated (confirm mastery before advancing), active-recall driven (you restate first; the agent fills gaps; quizzes via AskUserQuestion), and goal-bounded (the session does not end until your understanding is verified). Default subject = the current session/diff; pass a PR number, file, or subsystem to teach that instead. The pedagogical INVERSE of grill-me: grill-me stress-tests YOUR forward plan; comprehend transfers mastery of EXISTING work to you. Distinct from handoff (agent→agent narrative) and Bridge/Bookkeeping (agent→KG). Use when: (1) you want to deeply understand what was just built, (2) onboarding yourself or a teammate onto a PR / file / subsystem, (3) post-mortem learning after
broomva/skills · ★ 3 · AI & Automation · score 72
Install: claude install-skill broomva/skills
# comprehend — teach-to-mastery loop **Make the human deeply understand the work — verified, not assumed.** Distilled from Thariq Shihipar's teaching prompt (`research/entities/tool/thariq-teach-to-mastery-prompt.md`). This is a **composition skill**: it fires native primitives (`AskUserQuestion`, `/goal` P19, a persisted checklist doc P6) and adds only pedagogical sequencing on top — no infrastructure reimplemented. **Interactive-only.** comprehend depends on a live human to restate, answer quizzes, and signal mastery. In an autonomous, background, or piped run with no human responder, it must **abort** — never self-mark checklist boxes or self-certify understanding. The verification signal is the human's answer; without a human there is no signal. ## The one objective Understanding is the success metric. A summary the human nodded along to is a failure. The skill is complete only when the human has *demonstrated* mastery of every item on the checklist — where "demonstrated" means a correct, explained answer to a quiz, **not** the agent's own assessment that it taught well. ## Subject resolution | Invocation | Subject taught | |---|---| | `/comprehend` (bare) | The **current session / diff** — what we just did (recent `git diff`, the changes from this session). | | `/comprehend PR <n>` | The diff + discussion of pull request `n`. | | `/comprehend <path>` | A file or directory — its role, contracts, edge cases. | | `/comprehend <topic>` | A subsystem / concept already