agent-style

Literature-backed English technical-prose writing rules (agent-style, 21 rules). Its lane is FORMAL technical prose (papers, design docs, proposals, READMEs, commit messages). Use ONLY when the user explicitly asks for agent-style by name (e.g. "apply agent-style", "write/revise this per agent-style rules", "iterate until agent-style clean"). NOT for de-AI-ing casual or voiced text — use remove-ai-patterns for that. Do not auto-trigger for ordinary prose or documentation tasks.

Quality

81%

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

Quality

Content

72%

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

The body is well-structured for progressive disclosure — a tight rule overview pointing one level deep to real reference files — with concrete, actionable directives. The main gaps are an implicit rather than explicit iterate-until-clean feedback loop and some maintainer/meta sections that pad the token budget without aiding invocation.

Suggestions

Make the 'iterate until agent-style clean' loop explicit: add a short apply→re-scan-against-the-21-rules→fix sequence with a checkpoint so workflow_clarity reaches the top anchor.

Trim or relocate the 'Updating' and 'Optional Deterministic Audit' sections out of the invocation-time body (e.g. into a MAINTENANCE.md reference) to improve conciseness.

Fix the dangling reference to scripts/update-upstream.sh — no scripts/ directory exists in the bundle, so this path is currently broken.

Dimension	Reasoning	Score
Conciseness	The 21-rule list is dense and earns its tokens, but the verbatim Self-Verification Handshake response and the maintainer-oriented 'Updating'/'Optional Deterministic Audit' sections add material not needed at invocation time; the body could be tightened without losing clarity.	2 / 3
Actionability	Each rule is a concrete command-form directive and most carry inline example phrases to cut or avoid ('in order to', 'pushes the boundaries', 'leverage'), with a concrete 3-source resolution order for full rule bodies; per the scoring notes, absence of code in an instruction-only skill is not penalized when guidance is this specific.	3 / 3
Workflow Clarity	The resolution order for locating full rule bodies is a clear 3-step sequence and the rule list doubles as a checklist, but the 'iterate until agent-style clean' trigger implies an apply→verify→fix loop that is never spelled out with explicit checkpoints, so validation is implicit rather than staged.	2 / 3
Progressive Disclosure	SKILL.md is a compact overview with the 21 directives inline and full BAD/GOOD bodies deferred one level deep to references/RULES.md, clearly signaled via the resolution order; the referenced files (RULES.md, UPSTREAM-PIN, NOTICE.md, LICENSES/) all exist and are well organized.	3 / 3
	Total	10 / 12 Passed

Description

90%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it crisply scopes the lane, gives natural by-name trigger phrases, and explicitly disambiguates from the adjacent remove-ai-patterns skill. The only soft spot is specificity, since the capability is a single apply/revise action rather than a list of discrete operations.

Dimension	Reasoning	Score
Specificity	It names the domain concretely ('FORMAL technical prose (papers, design docs, proposals, READMEs, commit messages)') and gives action verbs in trigger examples ('apply', 'write/revise', 'iterate'), but 'what it does' is essentially one operation — apply a writing ruleset — rather than multiple discrete concrete actions, so it sits at 'names domain and some actions' rather than the multi-action top anchor.	2 / 3
Completeness	It answers both 'what' ('Literature-backed English technical-prose writing rules (agent-style, 21 rules)' plus its lane) and 'when' ('Use ONLY when the user explicitly asks for agent-style by name') with explicit triggers and exclusions.	3 / 3
Trigger Term Quality	Quotes natural phrasings a user would actually say — 'apply agent-style', 'write/revise this per agent-style rules', 'iterate until agent-style clean' — giving good coverage of how the skill is invoked by name.	3 / 3
Distinctiveness Conflict Risk	The 'Use ONLY when the user explicitly asks for agent-style by name' clause, the explicit hand-off to 'remove-ai-patterns' for casual/voiced text, and 'Do not auto-trigger for ordinary prose' give it a clear niche unlikely to fire for the wrong skill.	3 / 3
	Total	11 / 12 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 1 missing, 1 deeper-than-1-level	Warning

	Total	15 / 16 Passed

Repository: pchalasani/claude-code-tools
Path: plugins/writing/skills/agent-style/SKILL.md
Commit: 45ee0e2

Reviewed: 2 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.