putio/device-harness
Design small, typed, agent-friendly harnesses for hard-to-automate device and TV platforms. Use when shaping tools for Tizen, Roku, Android TV, Apple TV/tvOS, emulator wrappers, screenshot collectors, remote-control drivers, runtime assertion tools, or similar proof loops.
72
90%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong, well-crafted description that clearly defines a narrow, distinctive niche (harness patterns for device/TV platforms), lists concrete actions and platform targets, and includes an explicit 'Use when' clause with rich trigger terms. The boundary-setting ('Skip app-specific feature logic; this skill defines the reusable harness pattern language') further strengthens distinctiveness and reduces conflict risk.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and artifacts: 'typed, agent-friendly harnesses', 'emulator wrappers', 'screenshot collectors', 'remote-control drivers', 'runtime assertion tools', 'proof loops'. Also explicitly scopes out what it doesn't do ('Skip app-specific feature logic'). | 3 / 3 |
Completeness | Clearly answers both 'what' (design small, typed, agent-friendly harnesses for hard-to-automate platforms) and 'when' (explicit 'Use when' clause listing specific trigger scenarios like shaping tools for Tizen, Roku, Android TV, etc.). Also clarifies boundaries with 'Skip app-specific feature logic'. | 3 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'Tizen', 'Roku', 'Android TV', 'Apple TV', 'tvOS', 'emulator wrappers', 'screenshot collectors', 'remote-control drivers', 'harness'. These cover the specific platforms and tool types a user would mention when needing this skill. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche: device/TV platform test harness pattern design. The explicit scoping to 'reusable harness pattern language' and exclusion of 'app-specific feature logic' clearly differentiates it from platform-specific skills or general testing skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured harness design skill with strong actionability through concrete TypeScript types, CLI examples, and a clear seven-layer checklist. The workflow includes proper validation checkpoints and feedback loops. The main weaknesses are slightly verbose phrasing in places and the absence of the referenced bundle files (harness-pattern.md, examples.md, platform-notes.md) which are central to the workflow.
Suggestions
Provide the referenced bundle files (references/harness-pattern.md, references/examples.md, references/platform-notes.md) so the progressive disclosure structure actually functions as designed.
Tighten the opening paragraph — the description of what a harness 'should' do is redundant given the workflow and examples that follow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient but has some unnecessary elaboration. Phrases like 'The harness should wrap platform tools, expose a typed command surface, drive runtime behavior, assert meaningful state, and leave proof artifacts' restate what the workflow already shows. The guardrails section is mostly good but slightly verbose. The TypeScript types and CLI examples earn their place. | 2 / 3 |
Actionability | Provides concrete CLI command examples, a fully typed TypeScript interface with specific types, a clear output shape checklist, and specific guardrails. The seven-layer checklist from the workflow gives Claude a concrete verification step. The command surface and type definitions are copy-paste ready. | 3 / 3 |
Workflow Clarity | The 8-step workflow is clearly sequenced with explicit validation checkpoints: step 7 requires checking all seven layers before implementation, and step 8 is a feedback loop requiring revision if layers are missing, with a fallback strategy for unqueryable platform state. This is a well-structured design workflow with error recovery. | 3 / 3 |
Progressive Disclosure | References to three separate files (harness-pattern.md, examples.md, platform-notes.md) are well-signaled and one level deep, which is good structure. However, no bundle files were provided, meaning the referenced files don't actually exist in the evaluated bundle, making the references unverifiable. The SKILL.md itself contains a reasonable amount of inline content without being monolithic. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
Reviewed
Table of Contents