he-improve
Improve existing Harness Engineering implementations or workflows with evidence-backed changes. Use when users ask for targeted enhancement of shipped or drafted work.
32
26%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./Plugins/harness-engineering/fixtures/budget-archive/2026-04-21/deferred-store/skills/team_automation/he-improve/SKILL.mdQuality
Discovery
17%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 description is too vague and abstract to be useful for skill selection. It fails to explain what 'Harness Engineering' is, what concrete actions the skill performs, or what natural user requests should trigger it. The buzzword-heavy language ('evidence-backed changes', 'targeted enhancement') adds no discriminative value.
Suggestions
Define what 'Harness Engineering' means and list specific concrete actions the skill performs (e.g., 'optimize CI/CD pipelines, refactor deployment configurations, improve test harness coverage').
Add natural trigger terms users would actually say, such as specific technologies, file types, or task names (e.g., 'Harness CI/CD, pipeline optimization, .harness config files, deployment workflows').
Make the 'Use when...' clause more explicit with concrete scenarios, e.g., 'Use when users want to optimize Harness pipeline YAML, reduce build times, or refactor deployment stages.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague language like 'improve existing implementations or workflows' and 'evidence-backed changes' without listing any concrete actions. It doesn't specify what kinds of improvements, what 'Harness Engineering' entails, or what specific operations are performed. | 1 / 3 |
Completeness | It has a weak 'what' (improve implementations/workflows) and does include a 'Use when...' clause ('when users ask for targeted enhancement of shipped or drafted work'), but both are vague and lack explicit, actionable triggers. | 2 / 3 |
Trigger Term Quality | The terms 'Harness Engineering implementations', 'evidence-backed changes', and 'shipped or drafted work' are not natural phrases users would say. There are no common trigger words a user would naturally use when needing this skill. | 1 / 3 |
Distinctiveness Conflict Risk | The description is very generic — 'improve existing implementations or workflows' could apply to virtually any code improvement, refactoring, or optimization skill. Without specifics about what 'Harness Engineering' means or what domain it covers, this would easily conflict with other improvement/enhancement skills. | 1 / 3 |
Total | 5 / 12 Passed |
Implementation
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads as a high-level process framework rather than actionable guidance. It establishes good philosophical principles (evidence-based optimization, bounded experiments, fail-fast gates) and identifies useful anti-patterns, but fails to provide any concrete, executable steps — no commands, no code, no specific tool invocations. The domain-specific terminology ('Harness Engineering', 'context-disposition policy', 'measurement harness') is used without definition, making the skill opaque to anyone not already deeply familiar with the system.
Suggestions
Add concrete, executable examples for key procedure steps — e.g., show the actual commands to create an optimization spec, run a measurement baseline, and compare experiment results.
Replace the natural language 'Examples' section with at least one complete input/output walkthrough showing the spec format, measurement output, and decision outcome.
Define or link to definitions of domain-specific terms like 'Harness Engineering', 'measurement harness', 'context-disposition policy', and 'session-collector' so the skill is self-contained.
Fix the deeply nested relative path (../../../../../../references/session-evidence-contract.md) — either include the referenced file in the bundle or use a more stable path convention.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is moderately efficient but includes some unnecessary padding. Phrases like 'Progressive Disclosure Entry' and 'This entrypoint stays concise and keeps full operational context in archived references' are meta-commentary. The procedure steps are reasonably tight but use domain jargon ('Harness Engineering', 'session-collector', 'context-disposition policy') without concrete definitions, creating verbosity without clarity. Some sections like 'Full Context' with just icon references add little value. | 2 / 3 |
Actionability | The skill is almost entirely abstract and descriptive. There are no concrete commands, executable code, specific file paths to edit, or copy-paste ready examples. The procedure steps describe what to do conceptually ('Load or create the optimization spec', 'Run bounded iterations') but never show how. The 'Examples' section contains natural language prompts rather than input/output demonstrations with concrete artifacts. | 1 / 3 |
Workflow Clarity | The procedure provides a numbered sequence with validation checkpoints and the validation section includes explicit gates ('Fail fast: stop at first failed gate'). However, the steps are abstract rather than concrete — there are no specific commands, tools, or file operations shown. The feedback loop concept (keep/revise/discard based on measured outcomes) is present but implicit rather than demonstrated with concrete validation commands. | 2 / 3 |
Progressive Disclosure | The skill references an external file (session-evidence-contract.md) with a relative path, suggesting some progressive disclosure structure. However, no bundle files were provided to verify the reference exists, the path uses deeply nested relative navigation (../../../../../../), and the 'Full Context' section only links to icon assets rather than substantive reference materials. The skill claims to keep 'full operational context in archived references' but doesn't clearly signal where those references are. | 2 / 3 |
Total | 7 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_version | 'metadata.version' is missing | Warning |
Total | 10 / 11 Passed | |
- Repository
- jscraik/Agent-Skills
- Commit
4c78f98
- 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.