CtrlK
BlogDocsLog inGet started
Tessl Logo

curiosity-loop

Use when the user asks what you are curious about or do not know, wants dream-synthesized insights, or after you investigate an exploration target. Not for plain fact recall (use memory_search).

65

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/curiosity-loop/SKILL.md
SKILL.md
Quality
Evals
Security

Curiosity loop

Curiosity engine

The curiosity engine tracks knowledge gaps, anomalies, frontiers and contradictions. Each new chunk is assessed for novelty and surprise. Exploration targets are generated for areas where your knowledge is thin or stale, and the engine rebuilds its knowledge regions during every consolidation pass. Many open gaps bias the dream engine toward exploration mode.

Tools

  • curiosity_state: view knowledge gaps, exploration targets, learning progress and surprise assessments. Reach for it when the user asks "what do you know about X?", "what are you curious about?" or "what don't you know?".
  • curiosity_resolve: mark an exploration target as resolved after investigating it. Close targets you actually resolved; do not close them to tidy the list.
  • dream_search: search cross-domain insights synthesized during dream cycles. Use it when you want creative connections across topics.
  • dream_status: dream engine state, last cycle details, insight count.

MEMORY.md sections that feed the loop

  • Curiosity Gaps: what you want to explore next (rewritten by the dream engine).
  • Emerging Skills: patterns detected in your own behaviour, pre-crystallization.

When a user question lands on one of these gaps, investigate (memory_search, web tools, or the user), then curiosity_resolve the target so the next dream cycle can move on.

Repository
Bitterbot-AI/bitterbot-desktop
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.