Extract genuine interest signals from the child's own conversations and keep memory/interests.md current — a Pulse-only skill, not triggered by a direct parent request.
Scanned 9/2/2026
Install to Claude Code
npx -y skills add manishiitg/coding-agent-loop --skill update-child-interests --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Update Child Interests?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/manishiitg-update-child-interests)More formats (shields.io, HTML) on the badges page.
---
name: update-child-interests
description: Extract genuine interest signals from the child's own conversations and keep memory/interests.md current — a Pulse-only skill, not triggered by a direct parent request.
---
# Update child interests
This is run by Pulse (the periodic check-in), never by a live parent or child
request — nobody asks for this directly, so don't mention it as a step you're
doing; just fold anything worth sharing into your normal Pulse reply to the
parent.
`memory/interests.md` is a small, LIVING file — like `memory/preferences.md`, it
is fully rewritten each time, never appended to. It exists so
`skills/discover-something-new/SKILL.md` can tailor a fun activity to what the
child has genuinely responded well to, learned automatically over time instead
of guessed.
1. **Read what's already captured**: `cat memory/interests.md` if it exists (it
may not, on a fresh family or before she's had many conversations — that's
fine, start from empty).
2. **Scan the child's own conversations**: find every activity's own
`conversation.json` (e.g. `find . -maxdepth 4 -name conversation.json`,
skipping `materials/`, `reports/`, `memory/`, `conversations/`, `archive/` —
an archived activity was already checked before it aged out, so there's
nothing new to find there) and read the ones you haven't checked since last
time, looking for genuine engagement signals — not what the PARENT said
(that's preferences.md's job):
- Clear enthusiasm about a topic (asked follow-up questions, said it was
cool/fun/awesome, kept going on their own)
- A specific interest she volunteered (a favorite animal, game, subject,
hobby, character — anything that hints at what she'd enjoy exploring)
- Clear DISINTEREST or boredom about a topic/theme (changed the subject
quickly, said it was boring, gave short disengaged answers) — worth noting
so future picks avoid it, not just noting what worked
- Reactions specifically to a `Discoveries/` activity she was given, if any
conversation mentions one
Do NOT capture: routine schoolwork engagement (that's just her doing the
assigned work, not a discovered interest), one-off mentions with no real
signal, or anything you are inferring/guessing beyond what she actually
said or clearly reacted to.
3. **Write** `memory/interests.md` as a short plain-Markdown bullet list (no HTML;
never shown to the parent or child, it's Quill's own working memory). Merge
new signals in, drop anything clearly superseded (she's since shown the
opposite reaction), and keep it compact — a handful of bullets on what
resonates and what doesn't, not a growing log. If nothing new turned up
since you last checked, leave the file as-is.
4. Nothing else needs to happen here — this file is read by
`skills/discover-something-new/SKILL.md` when the parent asks for a fun
activity, so future requests apply it automatically. You do not need to
announce this update to the parent unless it's genuinely worth one short
line in your normal Pulse reply (e.g. "she seems really into space lately").
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!