scienceworld-task-interpreter

This skill parses a user's high-level scientific task in the ScienceWorld environment and extracts the core objective and target location. It should be triggered when a new task instruction is received, especially those involving finding, comparing, or manipulating objects. The skill interprets the query to identify the goal (e.g., 'find the animal with the shortest life span') and any specified locations (e.g., 'animals are in the outside location'), outputting a clear, actionable sub-goal for navigation or observation.

Quality

73%

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 ./experiments/src/skills/scienceworld/scienceworld-task-interpreter/SKILL.md

Quality

Content

63%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.

The content is reasonably concise and lays out a clear multi-step workflow with concrete ScienceWorld commands, but it is undermined by an empty 'Output Format:' section and an unreferenced glossary bundle file.

Suggestions

Complete the dangling 'Output Format:' section with a concrete Thought/output template (e.g., a worked example showing parsed objective, target, location, and resulting plan).

Link the existing references/action_glossary.md from the body (e.g., 'See [action_glossary.md](references/action_glossary.md) for the full command set') so the bundle file is discoverable.

Add a brief validation/feedback note in the Analyze & Execute step for cases where the expected object is not found at the location (e.g., re-observe or re-parse).

Dimension	Reasoning	Score
Conciseness	The body is lean and uses concrete commands (teleport, look around, focus on, pick up) without over-explaining concepts Claude already knows, though a few phrases like 'Read the user's instruction carefully' and 'This confirms the skill has correctly parsed the task' could be trimmed.	4 / 5
Actionability	Provides concrete executable commands for navigation and observation, but the 'Output Format:' section is left empty — a promised concrete template is missing entirely, leaving a key detail gap.	3 / 5
Workflow Clarity	The Parse → Formulate → Output sequence and the Navigate → Observe → Analyze & Execute sub-plan are clearly ordered, with a Thought step acting as a checkpoint, but there is no error-recovery feedback loop to reach a 5.	4 / 5
Progressive Disclosure	The body is organized into three clear sections, but the provided bundle file references/action_glossary.md is never linked or signaled from the SKILL.md body, so navigation to it is missing despite it being a real reference file.	3 / 5
	Total	14 / 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 well-crafted description that clearly communicates purpose and trigger conditions in third person with concrete examples. It is concise, actionable, and explicitly bounded to the ScienceWorld domain.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('parses', 'extracts the core objective and target location', 'interprets the query', 'outputting a clear, actionable sub-goal') but coverage is not fully comprehensive, sitting just below the anchor 5 example.	4 / 5
Completeness	Explicitly answers both what ('parses... extracts the core objective and target location', 'outputting a clear, actionable sub-goal') and when ('triggered when a new task instruction is received, especially those involving finding, comparing, or manipulating objects') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural trigger phrases ('new task instruction', 'finding, comparing, or manipulating objects') that a user would plausibly say, but lacks broader synonyms/variations that would warrant a 5.	4 / 5
Distinctiveness Conflict Risk	The ScienceWorld environment scoping gives it a clear niche, but the trigger verbs ('finding, comparing, manipulating objects') are somewhat generic, leaving minor overlap risk with other task-handling skills.	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-task-interpreter/SKILL.md
Commit: 076c6df

Reviewed: 6 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.