sandbox-failure-recovery

Recover from execute_code_sandbox failures by writing code to file and executing via run_shell

Quality

67%

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 ./benchmarks/gdpval/skills/sandbox-failure-recovery/SKILL.md

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 content is actionable with executable code and a clear validated workflow, but it is padded with a redundant complete example, a "Why This Works" section covering concepts Claude already knows, and a stray artifact, which hurts conciseness and progressive disclosure.

Suggestions

Remove the "Why This Works" section and the trailing `<THIS_SKILL_EVOLVED>` artifact to trim tokens Claude does not need.

Collapse the standalone "Complete Example" since it duplicates the step-by-step tool calls, or move it to a separate reference file.

Add an explicit "if verification fails, fix the script and re-run" feedback loop to strengthen the verify checkpoint.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient, but the "Why This Works" section restates facts Claude already knows, the complete example duplicates the step-by-step tool calls, and a stray `<THIS_SKILL_EVOLVED>` artifact adds noise.	2 / 3
Actionability	It provides complete, executable Python code and specific commands such as `python3 script_name.py` with concrete file paths, making the guidance copy-paste ready.	3 / 3
Workflow Clarity	A clear four-step sequence (preserve, write, execute, verify) is given with an explicit Step 4 "Verify Output" checkpoint, and the task is file-generation rather than destructive/batch so no feedback-loop cap applies.	3 / 3
Progressive Disclosure	No bundle files exist and sections are well-labeled, but everything is inline in a >50-line document with a redundant complete example that could be trimmed or split, so it is not a lean overview.	2 / 3
	Total	10 / 12 Passed

Description

57%

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 occupies a clear niche, but it lacks an explicit "Use when..." trigger clause and leans on technical jargon, leaving the when/trigger guidance implied rather than explicit.

Suggestions

Add an explicit "Use when execute_code_sandbox fails, times out, or the sandbox seems unstable..." clause to satisfy the completeness "when" requirement.

Soften the jargon by pairing the tool name with natural phrasing a user would say (e.g., "when sandbox code execution fails").

Optionally list one more concrete outcome (e.g., spreadsheet/file generation) to broaden the specificity beyond the two current actions.

Dimension	Reasoning	Score
Specificity	Names the domain and two concrete actions ("writing code to file" and "executing via run_shell"), but the action set is narrow and not comprehensive rather than a full list of specific capabilities.	2 / 3
Completeness	The "what" is explicit, but the "when" is only implied via the failure mention with no "Use when..." clause, which caps completeness at 2 per the guidelines.	2 / 3
Trigger Term Quality	"execute_code_sandbox failures" is a relevant trigger keyword but is technical jargon rather than natural user phrasing, and common variations are missing.	2 / 3
Distinctiveness Conflict Risk	The sandbox-failure-recovery niche with the distinct "execute_code_sandbox failures" trigger is clearly distinguishable and unlikely to fire for unrelated skills.	3 / 3
	Total	9 / 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/sandbox-failure-recovery/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.