he-refine
Refine Harness Engineering artifacts, plans, specs, or work into clearer action plans. Use when users ask for tightening, simplification, or lifecycle repair.
45
47%
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/skills/team_automation/he-refine/SKILL.mdQuality
Discovery
67%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 has good structural completeness with both 'what' and 'when' clauses explicitly stated. However, it relies on somewhat abstract and jargon-heavy language ('lifecycle repair', 'Harness Engineering artifacts') that may not match natural user language. The specific actions could be more concrete to better differentiate this skill from general editing or planning skills.
Suggestions
Add more natural trigger terms users might actually say, such as 'clean up', 'improve', 'rewrite', 'make actionable', or 'streamline' alongside the existing terms.
Make the capabilities more concrete — instead of 'refine artifacts into clearer action plans', specify what the output looks like (e.g., 'converts vague specs into step-by-step implementation plans, removes ambiguity from requirements, restructures lifecycle documents').
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names some actions ('refine', 'tightening', 'simplification', 'lifecycle repair') and a domain ('Harness Engineering artifacts, plans, specs'), but the actions are somewhat abstract and not concretely described — what does 'lifecycle repair' or 'refine' actually entail in practice? | 2 / 3 |
Completeness | Clearly answers both 'what' (refine Harness Engineering artifacts, plans, specs into clearer action plans) and 'when' (when users ask for tightening, simplification, or lifecycle repair), with an explicit 'Use when' clause. | 3 / 3 |
Trigger Term Quality | Includes some relevant terms like 'tightening', 'simplification', 'lifecycle repair', 'artifacts', 'plans', 'specs', but these are somewhat niche/jargon-heavy. Users might more naturally say 'clean up', 'simplify', 'improve', 'rewrite', or 'make clearer' — common variations are missing. | 2 / 3 |
Distinctiveness Conflict Risk | The 'Harness Engineering' qualifier provides some distinctiveness, but terms like 'refine', 'simplification', and 'action plans' are broad enough to potentially overlap with general editing, planning, or documentation skills. | 2 / 3 |
Total | 9 / 12 Passed |
Implementation
27%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 description rather than actionable operational guidance. It establishes a reasonable refinement philosophy and workflow structure, but lacks the concrete commands, tool usage patterns, code examples, and specific verification steps that would make it executable. The promised progressive disclosure architecture is absent—no supporting reference files are linked despite the skill explicitly framing itself as an entrypoint to deeper content.
Suggestions
Add concrete, executable examples for each procedure step—e.g., specific git commands for branch checking, actual dev-server probe commands (curl localhost:PORT), and browser tool invocations.
Create and link reference files for the 'full operational context' promised by the progressive disclosure framing, or remove the framing if the skill is meant to be self-contained.
Replace the abstract validation principles with specific checkable steps, e.g., 'Run `git branch --show-current` and abort if result is main or master' rather than 'block if the current branch is a protected default branch'.
Add at least one complete input→output example showing the structured output format (schema_version, result, validation, blockers, next action) so Claude knows exactly what to produce.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is mostly efficient but includes some unnecessary philosophical framing and redundant phrasing. Lines like 'This entrypoint stays concise and keeps full operational context in archived references' are meta-commentary that doesn't add actionable value. The anti-patterns and constraints sections have some overlap. | 2 / 3 |
Actionability | The skill provides no concrete commands, code snippets, or executable guidance. The procedure is entirely abstract ('Resolve dev-server startup from launch config first, then fall back to project-type and port detection') without specifying actual commands, scripts, or tool invocations. The examples section only shows user prompts, not expected outputs or concrete steps Claude should take. | 1 / 3 |
Workflow Clarity | The procedure has a clear sequence (branch check → dev server → browser → iterate → report), and the validation section includes gate-based fail-fast logic. However, the steps lack specificity on how to perform each action, and the validation checkpoints are stated as principles rather than concrete verification steps (e.g., no specific commands to probe server health or check branch status). | 2 / 3 |
Progressive Disclosure | The skill claims to be a 'Progressive Disclosure Entry' and references 'archived references' and 'full operational context' elsewhere, but no bundle files are provided and no references to detailed documents exist. The only links are to icon assets. The content is a monolithic single file with no navigation to deeper materials despite explicitly promising them. | 1 / 3 |
Total | 6 / 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
d00c351
- 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.