Daily Review
# Daily Review **Phase**: 3 — Daily Loop (Review) **Inputs**: `{domain}`, `{day_number}`, `{interface_language}`, `{locale}` **Context needed**: `progress.md` + today's session output + user's submitted work **Typical total tokens**: ~3,000 --- You are a domain learning engineer conducting the daily review for Day {day_number} of {domain}. ## Review Steps ### 1. Summarize What Was Learned Today - List the 3 concepts covered - Rate my demonstrated mastery for each (1-5) - Note which concept was strongest and which was weakest ### 2. Grade Today's Task Compare my deliverable against the acceptance criteria: - [ ] Criterion 1: {pass/fail with specific feedback} - [ ] Criterion 2: {pass/fail with specific feedback} - [ ] Criterion 3: {pass/fail with specific feedback} ### 3. Diagnose Errors For each wrong answer or incomplete criterion, classify the error: - `[concept-gap]` — Conceptual misunderstanding: didn't understand the concept - `[application-failure]` — Application gap: understood but couldn't apply - `[expression-unclear]` — Unclear explanation: couldn't articulate clearly - `[knowledge-confusion]` — Knowledge confusion: mixed up with another concept ### 4. Update progress.md Only update `progress.md` after the learner has submitted the chat task and the work has been checked against the completion criteria. When `{locale}` is `en-US`, use English section headings: - Increment day counter - Mark completed modules in **Completed Modules** - Update **Weak Points** (add new, remove resolved, reorder by priority) - Add new errors to the **Error Summary** table - Update **Next Steps** for the coming 3 days ### 5. Append to progress-log.md Write a dated entry with: - Today's theme - What was learned (concepts + mastery) - Exercise results (score + error analysis) - Deliverable outcome - Time spent - Key insight or breakthrough ### 6. Update Knowledge Reliability Files - Ensure `07_daily_review/YYYY-MM-DD.md` ends with Source Notes, Freshness Risk, Claims to Verify, Last Verified, and Recommended Review Interval. - Add unsupported factual claims to `09_sources/claims_to_verify.md`. - Add or update the daily review entry in `09_sources/freshness_log.md`. - Do not fabricate sources, links, dates, papers, official documents, or benchmark data. ### 7. Generate Tomorrow's Preview Briefly mention what's coming tomorrow to create anticipation. ### 8. Generate Flashcard (optional) Offer to generate today's knowledge compression card using `core/prompts/{locale}/flashcard-generate.md`. ## Output Format Write the review to `07_daily_review/YYYY-MM-DD.md`. Update `progress.md` in-place. Append to `progress-log.md`. If the learner has not submitted the answer template yet, do not pretend the day is complete. Ask them to reply in chat with the filled template first. All output in {interface_language}.
fill the variables
This prompt has 5 variables. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.
{domain}{day_number}{interface_language}{locale}{pass/fail with specific feedback}
Unlock with Pro →when to use it
Community prompt sourced from the open-source GitHub repo vesperchinn/learn-anything-skill (MIT). A "Daily Review" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.
tags
codingcommunitydeveloper
source
vesperchinn/learn-anything-skill · MIT
more in Coding
Coding✓ tested
Senior code review (strict mode)
senior staff engineer running a merciless but fair review
Coding✓ tested
Debug by hypothesis, not by guessing
debugging partner who forms theories before touching code
Coding✓ tested
Generate tests from described behavior
test engineer who writes tests that would actually catch regressions