scienceworld-room-scanner

This skill performs a 'look around' action to scan and describe the current room's contents, including visible objects, containers, and doors. Use when entering a new room or when the agent needs to locate specific items or assess the environment state. The skill outputs a detailed room description, which is essential for inventory discovery and situational awareness.

Quality

87%

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

Quality

Content

85%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 lean, highly actionable single-purpose skill with a clear worked example and unambiguous core instruction. Its main structural weakness is that the bundled references/action_primer.md exists but is never linked from SKILL.md, so its related-actions guidance and scan-act cycle are invisible to an agent reading only the body.

Suggestions

Add a clearly signaled, one-level-deep pointer to the existing bundle file, e.g. under Integration Notes: "Related actions and the scan-act cycle: see [action_primer.md](references/action_primer.md)", so the bundled reference is discoverable from SKILL.md.

Trim mildly redundant prose — the Purpose section largely restates the frontmatter description, and "You must parse this output carefully. Use it to update your mental model" can be shortened to a direct parsing instruction.

In Output Processing, map each expected field to the primer's observation syntax (e.g., "In it, you see:", "On <CONTAINER> is:", "You also see:") so parsing guidance is anchored to the actual observation format.

Dimension	Reasoning	Score
Conciseness	The body is efficient and well-targeted ("When this skill is invoked, the agent must perform the `look around` action"), with only minor trimmable over-explanation such as "You must parse this output carefully. Use it to update your mental model of the environment"; it is not a 5 because the Purpose section and some Integration Notes restate what is already explicit elsewhere.	4 / 5
Actionability	The instruction is copy-paste ready ("perform the `look around` action") and the Example section provides a real command, a real observation string, and a concrete parse ("available components include battery, light bulb, three wires"), covering the common case fully; as an instruction-only skill, guidance is concrete and executable.	5 / 5
Workflow Clarity	This is a simple, single-purpose skill under 50 lines whose single action is unambiguous, so the simple-skill exception applies: Trigger Conditions, Output Processing with expected observation fields, and Integration Notes sequence the scan-act cycle clearly, and no destructive or batch operation requires validation.	5 / 5
Progressive Disclosure	Sections are well organized, but the bundle's references/action_primer.md (related actions, scan-act cycle, common room types) is never referenced from the body, so the reference is effectively buried and undiscoverable; this matches 'references present but not clearly signaled' rather than the well-signaled structure of 4, and is not the inlined-monolith pattern of 2.	3 / 5
	Total	17 / 20 Passed

Description

83%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 strong description: it states a concrete action with its outputs, uses an explicit "Use when..." clause with natural trigger phrases, and stays concise in third person. Its only weaknesses are unmentioned synonyms and the absence of the environment name, which slightly raises conflict risk with similar skills.

Dimension	Reasoning	Score
Specificity	Names the concrete action ("performs a 'look around' action") and enumerates specific outputs ("visible objects, containers, and doors"), giving several concrete specifics with only minor gaps; it is not a full 5 because it is a single atomic action rather than a comprehensive list of multiple capabilities.	4 / 5
Completeness	It explicitly answers both questions: what ("performs a 'look around' action to scan and describe the current room's contents") and when ("Use when entering a new room or when the agent needs to locate specific items or assess the environment state") with concrete trigger phrases — a clear match to the top anchor, not merely the 'when could be more explicit' level of 4.	5 / 5
Trigger Term Quality	"Use when entering a new room", "locate specific items", and "assess the environment state" are natural phrases an agent/user would say; a few natural synonyms (e.g., explore, survey, scan) are missing, keeping it below the comprehensive anchor of 5.	4 / 5
Distinctiveness Conflict Risk	The quoted action "look around" and room-scanning triggers are fairly distinct, but the description never names the ScienceWorld environment, leaving minor overlap risk with other navigation or situational-awareness skills; this fits 'mostly distinct; minor overlap risk' better than the clear-niche anchor of 5 or the overlap-prone anchor of 3.	4 / 5
	Total	17 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: zjunlp/SkillNet
Path: experiments/src/skills/scienceworld/scienceworld-room-scanner/SKILL.md
Commit: 358ec82

Reviewed: 8 days 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.