CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

runpod-gpu-cloud

Safely inspect and operate RunPod resources with runpodctl, live product data, and explicit approval before paid or destructive actions.

Quality

69%

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 ./backend/cli/skills/cloud-compute/runpod/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

67%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.

The body is a well-organized, safety-first overview with a clear sequenced operating policy and explicit validation/approval checkpoints. Its main weakness is actionability: concrete commands are intentionally deferred to live `runpodctl --help`, so there are few copy-paste-ready examples for common operations.

Suggestions

Add one or two concrete, copy-paste-ready read-only command examples (e.g. a representative `runpodctl` list/status invocation) alongside the deferred-help guidance to lift actionability.

Render the Sources of truth links as markdown links with descriptive labels for clearer navigation and signaling.

Tighten the credential-boundary section, which restates the 'never print or persist the key' / 'cannot create...' constraints in slightly redundant forms across bullets.

DimensionReasoningScore

Conciseness

The body is dense and largely avoids explaining concepts Claude already knows; it stays focused on RunPod-specific boundaries and policy. A few phrases restate policy redundantly across sections, but overall it is efficient.

4 / 5

Actionability

It provides a concrete CLI contract (`runpodctl --help`, `runpodctl user`) and instructs reliance on live docs and `provider_compute` calls, but actual executable commands are deliberately deferred to live help, leaving the common mutation cases without copy-paste-ready examples.

3 / 5

Workflow Clarity

The seven-step Operating policy is a clearly sequenced workflow with explicit approval and reconciliation checkpoints ('Get explicit approval before...', 'reconcile with a read-only list/status command', 'Never blindly retry'), including feedback-style recovery guidance, satisfying the destructive-operation validation requirement with only minor gaps.

4 / 5

Progressive Disclosure

The content is organized into well-signaled sections (credential boundary, CLI contract, Operating policy, SSH handoff, Sources of truth) with external links one level deep and no bundle files to over-structure; it is an overview rather than a monolith, though the Sources of truth links are bare URLs rather than linked references.

4 / 5

Total

15

/

20

Passed

Description

58%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.

The description is specific to RunPod and conveys a clear safety-oriented 'what', but it omits an explicit 'Use when...' trigger clause and offers limited natural keyword variations, capping completeness and trigger quality.

Suggestions

Add an explicit 'Use when...' clause naming natural trigger phrases (e.g. 'Use when the user asks to launch, inspect, or stop RunPod GPU pods, check GPU availability, or manage RunPod compute').

Broaden trigger terms with common variations and synonyms such as 'GPU pods', 'RunPod instances', and 'cloud GPUs'.

Consider naming the concrete resource operations (create/start/stop/terminate Pods) in the description to lift specificity toward 5.

DimensionReasoningScore

Specificity

Names the domain (RunPod resources) and several concrete actions ('inspect', 'operate', 'runpodctl', 'live product data', 'explicit approval before paid or destructive actions'), but the actions are framed at a safety-policy level rather than enumerating the concrete resource operations in the description itself.

4 / 5

Completeness

It states a clear 'what' (inspect/operate RunPod resources safely) but provides no explicit 'when' / 'Use when...' trigger clause, so completeness is capped at 3 per the guideline.

3 / 5

Trigger Term Quality

Includes relevant terms like 'RunPod', 'runpodctl', and 'Pods', but lacks natural variations and common synonyms a user would say (e.g. 'GPU pods', 'RunPod instances', 'cloud GPUs') and gives no file-extension or product-surface breadth.

3 / 5

Distinctiveness Conflict Risk

RunPod-specific naming and the runpodctl tool give it a clear niche with minimal overlap risk against generic cloud skills, though broader GPU-cloud skills could share some triggers.

4 / 5

Total

14

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

synthetic-sciences/openscience

Path

backend/cli/skills/cloud-compute/runpod/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.