humanizer

Humanize text: strip AI-isms and add real voice.

Quality

58%

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 ./skills/creative/humanizer/SKILL.md

Quality

Content

61%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 skill is well-organized with concrete before/after examples and a clear self-auditing workflow, but the single-file monolithic pattern catalog and minor redundancy keep conciseness and progressive disclosure in the mid range. Actionability and workflow clarity are strong thanks to explicit tool calls and a feedback loop.

Suggestions

Move the 34-pattern reference catalog into a separate references file (e.g. PATTERNS.md) and keep SKILL.md as an overview pointing to it, improving progressive disclosure and token budget.

Add an explicit validation checkpoint for file edits (e.g. confirm the patch applies cleanly and re-read the changed section before showing the diff) to close the workflow-clarity gap.

De-duplicate the 'Key insight' explanation that appears both in the intro and the Attribution section to tighten conciseness.

Dimension	Reasoning	Score
Conciseness	The body is mostly purposeful pattern documentation, but at ~630 lines it repeats the same before/after structure 34 times and includes some restatement (e.g. the 'Key insight' appears twice, near the top and again in Attribution), so it could be tightened without losing clarity.	3 / 5
Actionability	Concrete before/after examples and named tool calls (read_file, patch, write_file) give mostly executable guidance, with only minor gaps such as voice calibration being described procedurally rather than with a worked mini-example.	4 / 5
Workflow Clarity	The task and Process sections lay out a clear numbered sequence with a built-in self-audit feedback loop ('What makes the below so obviously AI generated?' then revise), though file-edit operations lack an explicit validation checkpoint before applying patches.	4 / 5
Progressive Disclosure	No bundle files exist, so all 34 patterns plus example prose are inlined into a single SKILL.md; structure is clear via headers but the large pattern catalog is content that could live in a separate reference file, putting it at the 'some structure but inlined' anchor.	3 / 5
	Total	14 / 20 Passed

Description

56%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 concise and uses natural trigger language, but it omits an explicit 'Use when' trigger clause and lists only a couple of high-level actions, leaving completeness and specificity mid-range. It is reasonably distinct from sibling skills.

Suggestions

Add an explicit 'Use when...' clause naming concrete trigger phrases (e.g. 'Use when the user asks to humanize, de-AI, de-slop, or un-ChatGPT a piece of text, or to make a draft sound less LLM-written').

Expand the action list slightly (e.g. 'Identify and rewrite AI writing patterns, preserve meaning, and match a target voice') to raise specificity toward comprehensive coverage.

Include a couple of common synonyms or the file/editing contexts (drafts, PR descriptions, docs) to strengthen trigger-term coverage.

Dimension	Reasoning	Score
Specificity	Names two concrete actions ('strip AI-isms', 'add real voice') but they are high-level verbs rather than a comprehensive enumeration of capabilities, matching the '1-2 concrete actions but not comprehensive' anchor.	3 / 5
Completeness	It states a clear 'what' (humanize text by stripping AI-isms and adding voice) but lacks any explicit 'Use when...' trigger clause, which the guidelines say should cap completeness at 3.	3 / 5
Trigger Term Quality	Includes natural user phrases ('humanize', 'AI-isms', 'real voice') that users would plausibly say, with good but not exhaustive synonym/extension coverage; a few common variants like 'de-slop' or 'de-AI' are missing from the description itself.	4 / 5
Distinctiveness Conflict Risk	'Humanize text' and 'strip AI-isms' carve a fairly distinct niche with low overlap risk against other skills, though the brevity leaves minor ambiguity versus general editing skills.	4 / 5
	Total	14 / 20 Passed

Validation

75%

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 — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (648 lines); consider splitting into references/ and linking	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

	Total	12 / 16 Passed

Repository: NousResearch/hermes-agent
Path: skills/creative/humanizer/SKILL.md
Commit: 48f69e5

Reviewed: about 11 hours 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.