resilient-spreadsheet-workflow

Resilient multi-step workflow for spreadsheet processing when execute_code_sandbox fails, using shell_agent exploration, file-based scripts, and verification steps

Quality

81%

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

77%

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

The body is actionable and clearly sequenced with validation and error-recovery loops, but it spends tokens explaining shell basics Claude already knows and keeps all reference-worthy material inline rather than splitting it into bundle files. Trimming the redundant explanations and externalizing the diagnostic/error tables would tighten it further.

Suggestions

Remove explanations of basics Claude already knows (e.g., the sentence on what 2>&1 does, commentary on what 'which python' checks) and keep only the commands.

Move the 'Common Error Patterns' table and the Step 4 diagnostic command list into a references/ file (e.g., DIAGNOSTICS.md) referenced from the body, improving progressive disclosure for this 110-line skill.

Replace placeholder fragments like '<filename>' and '<processing script>' in the complete example with concrete sample values so the example is fully copy-paste ready.

Dimension	Reasoning	Score
Conciseness	The body explains basics Claude already knows — 'The 2>&1 redirection ensures both stdout and stderr are captured together' and what 'which python' checks — so it is mostly efficient but carries unnecessary explanation, landing at score 2.	2 / 3
Actionability	It provides concrete, copy-paste-ready commands ('python process_sheet.py 2>&1', "python -c 'import pandas'", 'ls -la input.csv', 'file input.csv', 'head -5 input.csv') and a runnable script template, matching the score-3 anchor.	3 / 3
Workflow Clarity	A clear 5-step sequence includes an explicit validation checkpoint ('Verify Output Before Reading') and an error-recovery feedback loop ('If Step 3 fails, add targeted verification commands'), matching the score-3 anchor.	3 / 3
Progressive Disclosure	There are no bundle files and all content — including the diagnostic command table and common error patterns — sits inline in a ~110-line file, exceeding the 50-line simple-skill exception, so it scores 2 rather than 3.	2 / 3
	Total	10 / 12 Passed

Description

85%

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, complete, and well-scoped to a distinct fallback niche, but its trigger language leans on internal tooling jargon ('execute_code_sandbox') and misses common user-facing terms like CSV, XLSX, or Excel. Strengthening trigger-term coverage would round out an otherwise strong description.

Suggestions

Add natural user-facing terms such as 'CSV', 'XLSX', 'Excel', or 'spreadsheet files' alongside 'spreadsheet processing' to broaden trigger coverage.

Reframe the 'when' trigger so it reads as something a user would naturally say (e.g., 'Use when spreadsheet processing with execute_code_sandbox fails or returns unknown errors') rather than relying solely on the internal tool name.

Consider whether 'shell_agent exploration, file-based scripts, and verification steps' are capabilities users care about or implementation mechanics better suited to the body.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions — 'shell_agent exploration, file-based scripts, and verification steps' — rather than only naming a domain, matching the score-3 anchor.	3 / 3
Completeness	It explicitly states what the skill does ('workflow for spreadsheet processing ... using shell_agent exploration, file-based scripts, and verification steps') and when to use it ('when execute_code_sandbox fails'), satisfying both halves with an explicit trigger.	3 / 3
Trigger Term Quality	It includes the natural term 'spreadsheet processing' but omits common variations (CSV, XLSX, Excel) and relies on internal jargon ('execute_code_sandbox fails') as the trigger, so it lands at score 2 rather than 3.	2 / 3
Distinctiveness Conflict Risk	The narrow fallback niche — spreadsheet processing specifically when execute_code_sandbox fails — is a distinct trigger unlikely to fire for unrelated skills, matching the score-3 anchor.	3 / 3
	Total	11 / 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/resilient-spreadsheet-workflow/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.