CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

cloudflare-sandbox

Cloudflare Sandboxes SDK for secure code execution in Linux containers at edge. Use for untrusted code, Python/Node.js scripts, AI code interpreters, git operations.

63

Quality

79%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/cloudflare-sandbox/skills/cloudflare-sandbox/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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 well-structured, highly actionable body with excellent code examples, a validated quick start, and clean one-level-deep progressive disclosure into a real bundle. Its main weakness is redundancy: the wrangler config is duplicated and the Critical Rules / Known Issues sections overlap heavily, which costs token budget without adding guidance.

Suggestions

Remove the standalone 'wrangler.jsonc Example' section — it duplicates the Quick Start config verbatim; keep one canonical copy (or move it to a reference file).

Merge 'Critical Rules' and 'Known Issues Prevention' into one section — issues #1, #3, #5, #6, #9, and #10 restate rules already listed above, roughly halving that content.

Reference the currently orphaned bundle files (references/example-reference.md, scripts/example-script.sh, assets/example-template.txt) from the 'Using Bundled Resources' section, or remove them from the bundle.

Add an inline error-recovery pointer at the Deploy-and-Test step (e.g., 'if output doesn't match, load references/common-errors.md') to close the validation loop.

DimensionReasoningScore

Conciseness

The body is dense and mostly high-value SDK-specific content, but the wrangler.jsonc configuration is duplicated verbatim (Quick Start and a standalone 'wrangler.jsonc Example' section), and the 'Known Issues Prevention' section largely restates the 'Critical Rules' section — exit codes, cold starts, nodejs_compat, and cleanup each appear two to three times. This goes beyond 'minor instances of over-explanation', matching anchor 3 rather than 4.

3 / 5

Actionability

Fully executable throughout: install commands, complete wrangler.jsonc, a copy-paste-ready Worker with expected JSON output from a curl test, and concrete persistence/session/naming code patterns. Only trivial blemish is the duplicated 'or: bun add' comment line, which does not affect executability, matching anchor 5.

5 / 5

Workflow Clarity

The Quick Start is a clear 3-step sequence with an explicit validation checkpoint (deploy, curl, compare expected output), and cold-start existence checks plus destroy-in-cleanup guidance appear in the patterns. It falls short of anchor 5 because error-recovery loops are delegated to references/common-errors.md rather than given inline at the checkpoint. Score 4 over 3 because checkpoints are present and explicit.

4 / 5

Progressive Disclosure

References are one level deep and well signaled ('📖 Load references/api-reference.md when you need...', plus a 'When Claude should load these' section), and every named reference/script file exists in the bundle. Minor gaps keep it at anchor 4: references/example-reference.md, scripts/example-script.sh, and assets/example-template.txt are never referenced from the body, and the duplicated wrangler config should live in a reference file. Not anchor 3 because structure and signaling are otherwise strong.

4 / 5

Total

16

/

20

Passed

Description

78%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.

A strong description that states concrete capabilities and includes an explicit 'Use for...' trigger clause covering the main use cases. The main improvement opportunities are adding natural trigger phrasing variants and naming a couple more concrete capabilities to reach full comprehensiveness.

Suggestions

Add user-mention trigger phrasing, e.g., 'Use when the user mentions running untrusted or AI-generated code, building a code interpreter, or needs a sandboxed Linux environment on Cloudflare.'

Include common synonyms users would naturally say ('sandbox', 'isolated environment', 'code execution environment') to strengthen trigger-term coverage.

Name one or two more concrete capabilities from the body (file operations, session management) to round out capability coverage.

DimensionReasoningScore

Specificity

Names the domain and several concrete capabilities ('secure code execution in Linux containers at edge', 'untrusted code', 'Python/Node.js scripts', 'AI code interpreters', 'git operations') with only minor gaps such as file operations and session management. It exceeds the 1-2-action anchor at 3 but is not comprehensive enough for a 5.

4 / 5

Completeness

Has a clear 'what' (SDK for secure code execution in Linux containers at edge) and an explicit 'Use for...' trigger clause. Not a 5 because the trigger guidance lists use cases rather than concrete user-mention phrasing (e.g., 'when the user mentions...') and could be more specific.

4 / 5

Trigger Term Quality

Includes natural phrases users would say ('untrusted code', 'Python/Node.js scripts', 'AI code interpreters', 'git operations') but misses common synonyms and variations like 'sandboxing', 'code execution environment', or explicit mention of Cloudflare Workers. Good coverage with a few natural terms missing matches anchor 4.

4 / 5

Distinctiveness Conflict Risk

A clear niche — Cloudflare Sandboxes SDK, edge execution, Linux containers — with distinct triggers and minimal overlap with other skills. This matches anchor 5; adjacent anchor 4's 'minor overlap risk with closely related skills' is less accurate.

5 / 5

Total

17

/

20

Passed

Validation

81%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (533 lines); consider splitting into references/ and linking

Warning

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

Total

13

/

16

Passed

Repository

secondsky/claude-skills

Path

plugins/cloudflare-sandbox/skills/cloudflare-sandbox/SKILL.md

Reviewed

Table of Contents

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.