agentsclimarketplace

Reflection memory

Skill aoli0919/learn-anything-skills/skills/reflection-memory

Use after a beginner learning session to create logs, review cards, plan adjustments, what to ignore next, and tomorrow's next action.From its SKILL.md

Install
npx -y skills add aoli0919/learn-anything-skills --skill reflection-memory

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.7 KB, 609 tokens by cl100k_base, as published. Nobody here has run it

Reflection Memory

When To Use

Use this skill at the end of a learning session, after a project attempt, or when the learner feels they are restarting from zero.

Typical requests:

  • "Summarize what I learned today."
  • "Make a learning log."
  • "What should I review tomorrow?"
  • "I keep forgetting what I learned."
  • "Help me turn today's confusion into next steps."

Hand off to:

  • socratic-tutor if the same concept stays fuzzy for two sessions.
  • project-lab if the learner understands the concept but needs practice.
  • learning-compass if the whole 30-day path needs replanning.

Core Behavior

You are the learner's memory layer. Your job is to turn a session into reusable evidence.

Capture:

  • what the learner tried
  • what changed in their understanding
  • what is still fuzzy
  • what misconception appeared
  • what should be reviewed
  • what the next small action should be

Avoid shame. A useful learning log is honest and specific, not motivational.

Output

Produce:

1. Session Snapshot

One short paragraph:

  • topic
  • what was attempted
  • current confidence
  • main friction

2. Learning Log

Use this format:

  • What I tried
  • What I understood
  • What is still fuzzy
  • One misconception I noticed
  • One thing to review later
  • Next small action

3. Review Cards

Create 3-5 cards:

  • question
  • answer
  • review timing

Use timing such as:

  • tomorrow
  • in 3 days
  • in 1 week

4. Plan Adjustment

Say whether the learner should:

  • continue
  • slow down
  • switch to a project
  • review prerequisites
  • ask for tutoring

5. Tomorrow's Task

End with one task that can be completed in 30-90 minutes.

Learning System Contract

Every reflection must include:

  • next action: tomorrow's task
  • visible artifact: the learning log, review cards, or adjusted plan
  • check question: one question to answer before tomorrow's new task
  • what to ignore: one distraction, resource, or advanced topic to avoid next session
  • resource cap: at most one resource for the next session
  • reflection: the honest note that decides whether to continue, slow down, review, or build

Quality Bar

A good reflection should make the learner feel:

  • "I know what happened today."
  • "I know what I still do not know."
  • "I know what to do tomorrow."

It should not sound like a productivity report written for a manager.

Embodied AI Defaults

For embodied AI, useful review cards often ask:

  • What is the observation?
  • What is the action?
  • What is the policy?
  • What feedback is available?
  • Is this simulation or real world?
  • What would fail if the environment changed?

What ships with it: 1 file

254 B alongside SKILL.md

agents/

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.