CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

generate-sandbox-policy

Generate sandbox security policies from plain-language requirements and optional REST API documentation. Produces L4 or fine-grained L7 network policies and ordered network middleware configuration. Use for API access rules, middleware host selection, failure behavior, or built-in and operator-run middleware attachment. Trigger keywords - generate policy, create policy, update policy, change policy, sandbox policy, network policy, API policy, security policy, allow API, restrict API, network middleware, supervisor middleware.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

73%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.

Highly actionable and workflow-transparent content: concrete templates, a decision tree, and thorough validation checklists make it easy to execute correctly. Its weaknesses are verbosity (duplicated guidance, paragraph-length warning strings) and monolithic structure — no reference files exist, and the single referenced file (examples.md) is missing from the bundle.

Suggestions

Create the referenced examples.md (or remove the dead link), and split schema-level detail — middleware binding rules, allowed_ips/private-IP semantics, and the quick-reference pattern library — into references/ files linked from the body, keeping SKILL.md as an overview.

Tighten Step 6's breadth warnings to one-line messages; the current paragraph-length entries (especially the L4-only warning) duplicate content that belongs in a reference or is explained elsewhere.

Deduplicate guidance that appears in multiple places: TLS handling is covered in Step 4, the hard-errors checklist, and the schema warnings; the 'Additional Resources' section repeats links and facts already stated in Step 3.

DimensionReasoningScore

Conciseness

Mostly domain-dense and non-redundant with Claude's prior knowledge, but noticeably padded: the breadth-warning table embeds paragraph-length warning strings, TLS and glob-semantics guidance is repeated across Steps 3-6, and "Additional Resources" restates links and facts already given in Step 3. This is more than the 'minor instances' of the 4 anchor, so it lands on 3.

3 / 5

Actionability

Fully executable throughout: YAML templates for policies and middleware, a decision tree, a path-to-glob mapping table, six copy-paste-ready quick-reference patterns, and concrete commands like `openshell sandbox create --policy <path>` and `openshell policy get <sandbox> --full` cover the common cases.

5 / 5

Workflow Clarity

Eight clearly sequenced steps with an explicit clarification loop (batch questions, stop rules, anti-over-interrogation guidance) and a dedicated validation step with hard-error, schema-warning, and structural checklists that act as explicit checkpoints before output, plus a confirm-and-refine loop.

5 / 5

Progressive Disclosure

The bundle contains no reference files at all, and the body's only reference — "For translation examples from real API docs, see [examples.md](examples.md)" — points to a nonexistent file. Schema-level detail, middleware binding rules, and the pattern library are all inlined in a ~670-line SKILL.md. Good section headers exist, but per the anchor this is content that clearly belongs in separate files inlined into one document, with a broken reference on top.

2 / 5

Total

15

/

20

Passed

Description

96%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: concrete, third-person statement of capabilities with an explicit 'Use for...' clause and a comprehensive trigger keyword list. The only weakness is that several trigger terms (e.g., "generate policy", "security policy") are generic enough to overlap with other policy-generation skills.

DimensionReasoningScore

Specificity

"Generate sandbox security policies from plain-language requirements and optional REST API documentation. Produces L4 or fine-grained L7 network policies and ordered network middleware configuration" names multiple concrete actions and outputs (L4 policies, fine-grained L7 rules, ordered middleware config) with comprehensive coverage, matching the top anchor rather than the 'minor gaps' of a 4.

5 / 5

Completeness

It explicitly answers both 'what' (generates L4/L7 network policies and ordered middleware configuration from NL requirements or API docs) and 'when' ("Use for API access rules, middleware host selection, failure behavior... Trigger keywords - ...") with concrete trigger phrases, matching the 5 anchor exactly.

5 / 5

Trigger Term Quality

The trigger list "generate policy, create policy, update policy, change policy, sandbox policy, network policy, API policy, security policy, allow API, restrict API, network middleware, supervisor middleware" comprehensively covers natural verb variations and synonyms a user would actually say; no common phrasing is missing.

5 / 5

Distinctiveness Conflict Risk

"sandbox policy", "network middleware", and "supervisor middleware" establish a clear niche, but generic triggers like "generate policy", "create policy", and "security policy" could fire for IAM, OPA, or cloud-policy generation skills. This sits between the 5 anchor (minimal conflict risk) and 4 (minor overlap with closely related skills) — the overlap is more than minimal, so 4.

4 / 5

Total

19

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (679 lines); consider splitting into references/ and linking

Warning

relative_links

Relative link issues: 1 missing

Warning

Total

14

/

16

Passed

Repository

NVIDIA/OpenShell

Path

skills/generate-sandbox-policy/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.