he-compound

Run a bounded Harness Engineering lifecycle across multiple stages. Use when the user wants coordinated brainstorm, spec, plan, work, review, and fix flow rather than one isolated stage.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

Tight, well-organized orchestration guidance with explicit validation checkpoints and fail-fast feedback loops, held back by an underspecified mode-selection decision and references to asset files that do not exist.

Suggestions

Specify the mode-selection decision criteria explicitly — which artifact signals route to lifecycle orchestration vs. solved-problem capture vs. resume-from-stage — so the core routing step is actionable rather than implicit.

Remove or fulfill the broken asset references in the Full Context section (the assets/ directory is absent), or replace them with a real one-level-deep reference if detailed material exists.

Collapse the overlap between Constraints and Anti-patterns (e.g., duplicate-doc creation appears in both) to trim redundancy, or explicitly frame Anti-patterns as the negative mirror of Constraints.

Dimension	Reasoning	Score
Conciseness	The body is lean and directive across well-separated sections (Philosophy, Procedure, Validation, Constraints, Anti-patterns) with no padding or explanation of concepts Claude already knows; every section earns its place.	3 / 3
Actionability	It names concrete targets (docs/solutions/, skill-refactor, skillify, he-compound-refresh, the one-file-write contract, full-vs-compact-safe defaults), but the core "Select lifecycle mode using artifact-first evidence" step never specifies which evidence maps to which mode, leaving a key detail incomplete.	2 / 3
Workflow Clarity	A numbered Procedure (steps 1-9) is paired with a dedicated Validation section and an explicit fail-fast feedback loop ("stop at first failed gate and do not continue"), matching the anchor for clear sequence with explicit validation and error-recovery checkpoints.	3 / 3
Progressive Disclosure	The body is sectioned rather than monolithic, but the only file references — [icon-small.png](./assets/icon-small.png) and [icon-large.png](./assets/icon-large.png) under Full Context — point to an assets/ directory that does not exist, so the references are broken and not genuinely navigable.	2 / 3
	Total	10 / 12 Passed

Description

85%

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 complete and fairly distinctive with an explicit Use-when clause and concrete stage enumeration, but its trigger terms are weighed down by Harness Engineering jargon and lack common variations.

Suggestions

Soften the jargon in the trigger clause by adding natural user phrasings (e.g., "Use when the user wants to take a feature from idea to fix in one coordinated pass") so it matches what users actually say.

Lead with the concrete action verbs (brainstorm, spec, plan, build/work, review, fix) before naming the abstract "lifecycle" to sharpen specificity.

Dimension	Reasoning	Score
Specificity	It lists multiple concrete stages — "coordinated brainstorm, spec, plan, work, review, and fix flow" — matching the anchor for enumerating several specific actions rather than vague abstraction.	3 / 3
Completeness	It states what ("Run a bounded Harness Engineering lifecycle across multiple stages") and an explicit when ("Use when the user wants coordinated... flow rather than one isolated stage"), satisfying both halves with an explicit trigger.	3 / 3
Trigger Term Quality	The stage names (brainstorm, spec, plan, work, review, fix) are somewhat natural, but the framing leans heavily on internal jargon ("Harness Engineering lifecycle") and misses common user variations, so it is not a 3.	2 / 3
Distinctiveness Conflict Risk	The "rather than one isolated stage" trigger carves a clear compound-orchestration niche distinct from single-stage skills, making wrong-skill triggering unlikely; it is not a 2 because the differentiator is explicit rather than merely domain-specific.	3 / 3
	Total	11 / 12 Passed

Validation

87%

Warnings & errors only

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

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
relative_links	Relative link issues: 2 missing	Warning

	Total	14 / 16 Passed

Repository: jscraik/Agent-Skills
Commit: 5a6027f

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

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.