alfworld-object-picker

Picks up a specified object from a given receptacle. Use this skill when the agent has located a required object and needs to acquire it for later use, such as taking an item from a surface or container. The skill requires the object and source receptacle as inputs, executing a 'take' action to transfer the object into the agent's inventory, enabling further manipulation like placement or usage.

Quality

71%

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/alfworld/alfworld-object-picker/SKILL.md

Quality

Content

76%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-structured instruction-only skill for a single ALFWorld action: lean on tokens and concrete in its action format. Its main gap is workflow validation — the success/failure loop is delegated to a reference rather than handled inline with an explicit checkpoint.

Suggestions

Add an explicit validation checkpoint after 'take', e.g. confirm the observation reports the object now in inventory before proceeding, instead of only branching to troubleshooting on 'Nothing happened'.

Inline a one-line success check (observation contains 'You pick up') so the feedback loop is self-contained rather than deferred to the reference.

Optionally note expected failure modes inline (wrong location, exact name mismatch) before pointing to troubleshooting.md, so the core workflow stands alone.

Dimension	Reasoning	Score
Conciseness	The body is a lean ~15 lines with well-organized Inputs/Process/Output sections and no padding or explanation of concepts Claude already knows; every line earns its place, matching the 'lean and efficient' anchor.	5 / 5
Actionability	It gives the concrete action format 'take {object} from {source_receptacle}' with real example identifiers ('toiletpaper 1', 'toilet 1') and points to a troubleshooting reference, but it is a templated action string rather than copy-paste code covering multiple cases, so it sits at mostly-executable rather than fully-executable.	4 / 5
Workflow Clarity	Steps are clearly sequenced (Verify Context, Execute Action, Handle Feedback), but the destructive/batch cap applies: pickup mutates inventory and the flow defers failure handling to an external reference rather than validating success before proceeding, capping this at 3; it is not 4 because the validation checkpoint is implicit, not an explicit verify step.	3 / 5
Progressive Disclosure	The overview body is appropriately short and offloads failure detail to a real one-level-deep reference (references/troubleshooting.md, which exists), with clear signaling; it is not 5 because the single reference is the only detail file and navigation could be more explicitly structured, but for a skill of this size the split is good.	4 / 5
	Total	16 / 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 concrete, third-person, and answers both what and when with explicit inputs and outcome, placing it solidly above midpoint. It could be lifted toward the top anchors by adding natural user-facing trigger terms and a sharper 'when' clause tied to concrete user requests.

Suggestions

Add natural user-facing trigger terms ('pick up', 'grab', 'take an item from') so the 'when' clause matches phrases a user would actually say.

Tighten the 'when' clause to concrete triggers, e.g. 'Use when the user asks to pick up, grab, or take a specific item from a known receptacle in ALFWorld.'

Consider noting the ALFWorld domain explicitly to reduce conflict risk with generic object-handling skills.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete actions ('take' action, 'transfer the object into the agent's inventory', 'placement or usage'), with only minor gaps in coverage; it does not reach a 5 because the action set is essentially a single take operation rather than multiple distinct concrete actions.	4 / 5
Completeness	It explicitly answers both 'what' (picks up an object via a take action) and 'when' ('Use this skill when the agent has located a required object and needs to acquire it for later use'), plus inputs and outcome; the 'when' clause is present and reasonably specific but framed around agent state rather than concrete user triggers, so it stops short of 5.	4 / 5
Trigger Term Quality	It includes 'Use this skill when the agent has located a required object and needs to acquire it for later use' and concrete trigger nouns like 'surface or container', but lacks the natural short synonyms a user would say ('pick up', 'grab', 'take item') and is phrased from the agent's perspective rather than a user's; not a 4 because common variations are missing.	3 / 5
Distinctiveness Conflict Risk	The narrow scope (take object from a specific receptacle into inventory in the ALFWorld domain) makes it mostly distinct from sibling movement or placement skills, with only minor overlap risk against a generic 'place' skill; not a 5 because the trigger phrasing is not crisp enough to fully eliminate overlap.	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/alfworld/alfworld-object-picker/SKILL.md
Commit: cce11c3

Reviewed: 2 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.