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).

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/curiosity-loop/SKILL.md

Quality

Content

86%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, well-organized instruction-only skill body: domain-specific knowledge is delivered without padding, each tool has concrete trigger conditions, and the resolve-honestly guardrail shows real judgment about failure modes. The only meaningful gap is the absence of a worked example and a single consolidated statement of the loop, which keeps actionability and workflow clarity at 4.

Dimension	Reasoning	Score
Conciseness	At ~20 lean lines, every sentence carries domain-specific information Claude could not infer ("Many open gaps bias the dream engine toward exploration mode", "Close targets you actually resolved; do not close them to tidy the list") with zero padding or explanation of known concepts — a direct match for anchor 5.	5 / 5
Actionability	Per-tool guidance is concrete and conditional ("Reach for it when the user asks 'what are you curious about?'"), and the closing loop names exactly which channels to investigate through (memory_search, web tools, or the user). It falls short of anchor 5 because there are no worked examples of a gap entry, a resolution, or a dream insight — the common cases are described but not shown.	4 / 5
Workflow Clarity	The loop is sequenced (question lands on a gap → investigate → curiosity_resolve → next dream cycle) with a built-in honesty check ("do not close them to tidy the list"), but the sequence is distributed across the Tools section and the closing paragraph rather than presented as one explicit workflow, and when to consult dream_status versus dream_search is only implied. Solidly anchor 4, not 3 (checkpoints are present) nor 5 (no single end-to-end sequence).	4 / 5
Progressive Disclosure	Under 50 lines, single-purpose, no bundle files exist and none are needed; the three well-labeled sections (engine overview, tools, MEMORY.md sections) fully organize the content, satisfying the rubric's explicit simple-skill exception for a score of 5.	5 / 5
	Total	18 / 20 Passed

Description

73%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A well-scoped, trigger-forward description with an excellent explicit 'when' clause and a rare negative-boundary disambiguation against memory_search. Its main weakness is that the 'what' is never stated — the skill's actual capabilities (viewing gaps, resolving targets, searching dream insights) must be inferred from trigger phrases. Third-person voice is used correctly and there is no padding.

Suggestions

Open with a concrete capability statement, e.g. "Views knowledge gaps and exploration targets, resolves investigated targets, and searches cross-domain dream-synthesized insights. Use when..." so the 'what' is explicit rather than inferred.

Add the natural trigger phrasing from the body ("what do you know about X?") to the description to broaden trigger coverage beyond "curious" and "do not know".

Dimension	Reasoning	Score
Specificity	The description names its domain ("curiosity", "dream-synthesized insights", "exploration target") and gestures at 1-2 actions ("investigate an exploration target"), but never states what the skill actually does (view knowledge gaps, resolve targets, search synthesized insights) — those capabilities live only in the body. It sits above score 2 (domain named, actions generic) but below 4 because it lists no concrete capability.	3 / 5
Completeness	The 'when' is fully explicit ("Use when the user asks... or after you investigate an exploration target") and even adds a negative boundary ("Not for plain fact recall"), but the 'what' is only implied through triggers rather than explicitly stated — the mirror image of anchor 4's caveat, and clearly above anchor 3 where 'when' is weak or missing.	4 / 5
Trigger Term Quality	Phrases like "asks what you are curious about or do not know" and "wants dream-synthesized insights" are natural user utterances with good coverage, but common variants (e.g., "what do you know about X", "surprise me") are absent — matching anchor 4 rather than the comprehensive synonym coverage of anchor 5.	4 / 5
Distinctiveness Conflict Risk	The trigger set (curiosity, knowledge gaps, dream-synthesized insights, exploration targets) occupies a clear niche, and the explicit redirect "Not for plain fact recall (use memory_search)" preempts the one obvious conflict, matching anchor 5's minimal-conflict profile.	5 / 5
	Total	16 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: Bitterbot-AI/bitterbot-desktop
Path: skills/curiosity-loop/SKILL.md
Commit: 1c7dc1b

Reviewed: 19 hours ago

Table of Contents

Discovery Implementation Validation

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.