scienceworld-target-identifier

Analyzes room observations to identify objects matching a given target description (e.g., 'living thing'). Triggered after exploring a room when the agent needs to locate a specific type of item. Processes the observation list, filters objects based on target criteria, and returns candidate objects for further action.

Quality

65%

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-target-identifier/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 a well-structured, mostly lean instruction skill with a clear workflow and useful error-handling feedback loops. Its credibility is undermined by a reference to a bundled classification script that does not exist and an unreferenced, truncated action_patterns.md bundle file.

Suggestions

Either add the referenced classification script to the bundle or rewrite Step 2 to use the existing taxonomy.md directly with concrete filtering instructions.

Replace prose mentions of bundled resources with explicit markdown links (e.g. 'See [taxonomy.md](references/taxonomy.md)') and either reference or remove the incomplete action_patterns.md.

Add an explicit validation checkpoint in the workflow (e.g. verify each candidate's exact name against the observation before returning it) to lift workflow clarity toward 5.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence; environment-specific facts (the 'Here you see:' format, teleportation and open-container rules) genuinely add information Claude does not already know, with only minor passages that could be trimmed.	4 / 5
Actionability	Step 2 instructs to 'Use the bundled classification script to filter objects', but no such script exists in the bundle (only taxonomy.md and action_patterns.md), leaving a broken execution path despite the concrete worked example at the end.	3 / 5
Workflow Clarity	Steps 1–4 are clearly sequenced and the Error Handling section provides feedback loops (no match → teleport and repeat; ambiguous → examine; uncertain → check taxonomy), but the main workflow lacks an explicit validation checkpoint before returning candidates.	4 / 5
Progressive Disclosure	Sections are organized, but references are not surfaced as proper links (e.g. 'Check reference taxonomy in bundled resources' rather than a markdown link), one reference points to a non-existent classification script, and the bundled action_patterns.md is neither referenced in the body nor complete.	3 / 5
	Total	14 / 20 Passed

Description

67%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.

The description is specific and clearly scoped to the ScienceWorld environment, articulating concrete actions and an explicit trigger condition. Its main weakness is that the trigger is framed around agent state rather than natural user language, leaving trigger-term quality and completeness just below the top anchor.

Suggestions

Rewrite the trigger as a natural 'Use when...' clause a user would say, e.g. 'Use when you need to find a specific type of object (e.g. a living thing) after looking around a room.'

Add common synonyms and concrete trigger phrases (e.g. 'find a living thing', 'locate a container', 'identify an electrical device') to broaden natural keyword coverage.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions — 'Analyzes room observations to identify objects', 'filters objects based on target criteria', 'returns candidate objects' — with only minor coverage gaps, fitting the 4 anchor rather than the comprehensive 5.	4 / 5
Completeness	It explicitly answers both what (analyze/filter/return candidate objects) and when ('Triggered after exploring a room when the agent needs to locate a specific type of item'), but the when is conditional on internal agent state rather than a concrete user-facing trigger phrase, so it stops short of 5.	4 / 5
Trigger Term Quality	It includes a relevant trigger ('Triggered after exploring a room when the agent needs to locate a specific type of item') and the example 'living thing', but these are tied to agent state rather than natural user phrases, missing common variations — matching the 3 anchor.	3 / 5
Distinctiveness Conflict Risk	It carves a clear ScienceWorld niche ('Analyzes room observations to identify objects matching a given target description') with minimal overlap risk against unrelated skills, though it could still brush against other ScienceWorld observation-handling skills.	4 / 5
	Total	15 / 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-target-identifier/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.