agent-harness-refactor

Reestructura un arnes sin cambiar comportamiento, con disciplina deletion-only para compresion: insertions=0 verificable en el receipt, unidades borradas contadas y afirmaciones sobrevivientes intactas por construccion. Usar al pedir 'refactorizar arnes', 'comprimir AGENTS.md', 'deletion only', 'reordenar harness sin cambiar reglas'. NO para simplificar la topologia del sistema (agent-harness-simplify) ni reparar fallos (agent-harness-repair).

Quality

77%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/claude-native-toolkit/skills/agent-harness-refactor/SKILL.md

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 content is concise and well-structured with a clear procedure and validation step, but it lacks concrete executable commands and references some bundle directories that do not exist. Tightening the actionability and aligning the Packet listing with the real bundle would raise quality.

Suggestions

Add concrete, executable commands or code snippets for the core steps (e.g., the diff insertions=0 check and the harness_lib.structured_lines comparison) instead of only describing them.

Inline-define opaque terms like 'score' and 'Q9 en verde' or link to where they are defined, so the procedure is self-contained.

Align the Packet layer listing with the actual bundle contents, or mark the advertised directories (knowledge/, prompts/, examples/, agents/) as optional/expected-only.

Dimension	Reasoning	Score
Conciseness	The body is lean and avoids explaining concepts Claude already knows, though the generated wire-packet block describing layer directories adds some decorative padding that could be trimmed.	4 / 5
Actionability	Steps give specific direction ('Verifica insertions=0 en el diff', 'Receipt: unidades borradas, afirmaciones sobrevivientes') but lack copy-paste executable commands, and terms like 'score' and 'Q9 en verde' are referenced without being defined inline.	3 / 5
Workflow Clarity	A clear five-step sequence is present with a re-audit validation checkpoint (step 5), satisfying the destructive-operation requirement, though the retry path if the re-audit fails is not spelled out as an explicit feedback loop.	4 / 5
Progressive Disclosure	The body is a well-organized overview pointing one level deep to a real references/deletion-only.md, but the Packet section advertises directories (knowledge/, prompts/, examples/, agents/) that are not present in the actual bundle.	4 / 5
	Total	15 / 20 Passed

Description

87%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 strong: it states concrete capabilities, gives explicit trigger phrases, and clearly distinguishes itself from sibling skills. The main gap is slightly less concrete enumeration of the reorganization variant's actions.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete actions ('insertions=0 verificable en el receipt, unidades borradas contadas y afirmaciones sobrevivientes intactas por construccion'), but reorganization actions are only implied via trigger phrases rather than enumerated, leaving minor coverage gaps.	4 / 5
Completeness	Explicitly answers both what ('Reestructura un arnes sin cambiar comportamiento...') and when ('Usar al pedir...') with concrete trigger phrases plus explicit 'NO para...' exclusions.	5 / 5
Trigger Term Quality	Provides natural trigger phrases a user would say ('refactorizar arnes', 'comprimir AGENTS.md', 'deletion only', 'reordenar harness sin cambiar reglas') with synonyms (arnes/harness), though a few natural variations are absent.	4 / 5
Distinctiveness Conflict Risk	Clear niche with explicit boundaries against named sibling skills (agent-harness-simplify, agent-harness-repair), minimizing conflict risk.	5 / 5
	Total	18 / 20 Passed

Validation

68%

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 — 11 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 1 missing	Warning

	Total	11 / 16 Passed

Repository: JaviMontano/claude-plugins
Path: plugins/claude-native-toolkit/skills/agent-harness-refactor/SKILL.md
Commit: e8f986b

Reviewed: 3 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.