web-tool-fallback

Implement fallback strategies when web-reading tools fail simultaneously

Quality

62%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./benchmarks/gdpval/skills/web-tool-fallback/SKILL.md

Quality

Content

50%

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

The skill offers a clear, actionable fallback workflow with good error-recovery loops, but quality is undermined by duplicated headers and steps, a malformed code fence, and a monolithic structure with no progressive disclosure.

Suggestions

Remove the duplicated Step 2 and Step 2b headers, duplicated list items, and duplicated decision-tree lines so the sequence reads as a single clean path.

Fix the unclosed ```python fence around the embedded-knowledge bullet list (lines ~75-80), which currently renders prose as code and breaks the document structure.

Replace the pseudocode in 'Example Application' (execute_code_sandbox(), create_with_placeholders()) with concrete, executable commands, or clearly label it as illustrative pseudocode.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and assumes Claude's competence, but it is padded with duplicated headers (Step 2 and Step 2b appear twice), duplicated list items, and an explanatory 'Why run_shell works' section. Not a 3 due to this redundant padding; not a 1 because it does not explain basic concepts Claude already knows.	2 / 3
Actionability	It provides concrete, executable bash and Python examples (python3 -c, heredoc), but a broken/unclosed ```python fence around the embedded-knowledge list and pseudocode in the 'Example Application' (execute_code_sandbox(), create_with_placeholders()) make parts incomplete. Not a 3 because not everything is copy-paste ready; not a 1 because substantial executable guidance exists.	2 / 3
Workflow Clarity	A clear sequence with explicit feedback loops (sandbox fail -> immediately use run_shell, no retries) and a checklist exists, but duplicate Step 2 and Step 2b headers plus duplicated decision-tree lines muddy the actual sequence. Not a 3 because the duplication breaks sequence clarity; not a 1 because steps and checkpoints are clearly present.	2 / 3
Progressive Disclosure	The body is organized into clear sections (When to Use, Recognition Criteria, Decision Tree, Best Practices), but it is a ~200-line monolithic file with no bundle files and no reference split. Not a 3 because it exceeds the small-skill threshold without external reference separation; not a 1 because sections are well-organized rather than a wall of text.	2 / 3
	Total	8 / 12 Passed

Description

75%

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 clearly states both purpose and trigger condition with a distinctive, narrow activation scenario. It is slightly held back by an abstract single action verb and technical rather than natural trigger phrasing.

Dimension	Reasoning	Score
Specificity	The phrase 'Implement fallback strategies' names a domain (web tool failure) and a single action, but it is abstract and not a comprehensive list of concrete actions. It is not a 1 because an action is named, and not a 3 because no multiple specific concrete actions are enumerated.	2 / 3
Completeness	It explicitly answers both what ('Implement fallback strategies') and when ('when web-reading tools fail simultaneously'), providing an explicit trigger clause equivalent to 'Use when...'. Not a 2 because the 'when' trigger is explicit rather than implied.	3 / 3
Trigger Term Quality	It includes relevant keywords ('web-reading tools', 'fail simultaneously', 'fallback') but the phrasing is technical and a user would be unlikely to say 'web-reading tools fail simultaneously' verbatim. Missing common variations keeps it at 2 rather than 3.	2 / 3
Distinctiveness Conflict Risk	The trigger (all web-reading tools failing simultaneously) is a narrow, distinct niche unlikely to fire for other skills. Not a 2 because it is more specific than 'Works with document files' and has minimal overlap.	3 / 3
	Total	10 / 12 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: HKUDS/OpenSpace
Path: benchmarks/gdpval/skills/web-tool-fallback/SKILL.md
Commit: c5a9c4b

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.