2.5k
GitHub Stars
6
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill meditate- _meta.json451 B
- feedback.md1.7 KB
- memory-template.md1.6 KB
- sandbox.md1.4 KB
- SKILL.md4.3 KB
- topics.md2.9 KB
Overview
This skill helps an agent think proactively during idle time by producing sandboxed reflections, adaptive rhythms, and feedback-driven focus areas. It generates short, non-actionable observations and questions, queues them locally, and adapts to user reactions. The skill is strictly reflective: it never executes actions or modifies external systems.
How this skill works
The skill reads recent conversation history and memory files stored in ~/meditate/ to detect patterns and a likely user profile. It generates text-only insights formatted as observations and questions, stores up to three pending items in a local queue, and archives delivered items. Frequency and breadth of meditations adapt based on activity levels and explicit feedback.
When to use it
- Agent has idle time between user interactions and should think proactively.
- When you want insights, questions, or observations without executing actions.
- To surface strategic gaps, recurring patterns, or idea prompts for later review.
- When you want a low-interruption way to refine focus areas via feedback.
- To maintain reflective habit-generation without accessing external services.
Best practices
- Keep meditations short initially: 1–2 observations until feedback is received.
- Confirm detected profile changes before updating stored preferences.
- Respond with simple feedback (useful / not relevant / don’t think about X) to tune priorities.
- Limit meditation topics and never include personal data or external analyses.
- Treat every output as a prompt for thought—frame as questions, never action items.
Example use cases
- Nightly brief reflections that raise strategic questions for an entrepreneur’s projects.
- Periodic architecture observations for a developer that highlight potential technical debt patterns.
- Creative prompt nudges that surface style changes or recurring prompt structures.
- Habit and goal-oriented questions for personal users to re-evaluate priorities.
- Archiving delivered insights and using feedback to reorder meditation topics.
FAQ
No. It only generates text reflections and never executes commands or modifies files outside its local memory folder.
How often does it meditate?
Frequency adapts to user activity: nightly brief meditations for very active users, several times per week for moderate activity, and weekly for low activity. Silence reduces frequency.