Canon

Name: Canon
Author: AgriciDaniel

ASecurity

Ledger: 025 | source: blader humanizer agent skill | confidence: evidence-based | fold: [[blader humanizer agent skill]] | status: captured.

51 stars
0 votes
0 copies
2 views
Added September 19, 2026

ai-agentsgitdocumentation

Security analysis

A100/100

Pro scans all 21 files and shows the line behind each finding

Scanned September 19, 2026

npx -y skills add AgriciDaniel/anti-slop --skill canon --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Canon?

Add the live security badge to your README. It updates with every re-scan.

[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/agricidaniel-canon/badge)](https://www.skillsdirectory.com/skills/agricidaniel-canon)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro

SKILL.md

---
type: "canon"
title: "025. blader humanizer agent skill"
domain: "detection and repair of AI slop in prose, code, documentation, and agent output, grounded in corpus evidence rather than authorship detection"
status: "active"
created: "2026-07-28"
updated: "2026-07-28"
tags:
  - "#domain/detection-and-repair-of-ai-slop-in-prose-code-documentation-and"
  - "#type/canon"
  - "#confidence/evidence-based"
confidence: "evidence-based"
related:
  - "[[blader humanizer agent skill]]"
  - "[[Index]]"
  - "[[Dashboard]]"
  - "[[CONVENTIONS]]"
  - "[[Tag Taxonomy]]"
  - "[[Source Intake Workflow]]"
  - "[[Research Refresh Workflow]]"
  - "[[Claim Verification Flow]]"
  - "[[Synthesis Workflow]]"
  - "[[Reporting Workflow]]"
  - "[[Source Manifest Guide]]"
  - "[[Best Practices Kernel]]"
  - "[[Health Scorecard]]"
  - "[[Action Roadmap]]"
  - "[[Weekly Report]]"
  - "[[Approval Queue]]"
source_urls:
  - "https://github.com/blader/humanizer"
---

# 025. blader humanizer agent skill

Ledger: 025 | source: blader humanizer agent skill | confidence: evidence-based | fold: [[blader humanizer agent skill]] | status: captured.

## What It Says

- The main branch declares version 3.0.0 and an MIT license
- The skill now enumerates 25 patterns, with the first five permitting an edit on one sighting and weak-alone patterns requiring corroborating tells in the same passage
- Its rewrite instructions preserve supported claims and forbid adding unsupported facts, but this remains a prompt instruction rather than an external verification mechanism
- Section 8 bars em and en dashes unless the supplied writer sample uses them, while exempting code, commands, paths, and URLs
- File mode says to preserve code blocks, inline code, commands, paths, YAML metadata, data, and link targets

## Source

Source: [blader humanizer agent skill](https://github.com/blader/humanizer); type supporting; retrieved 2026-09-11; refresh_due 2026-10-11.

## Brain Hooks

- Folded concept: [[blader humanizer agent skill]]
- Source ledger: `references/source-ledger.json`
- Verification path: [[Claim Verification Flow]]