Diagnose, rewrite, and validate text to reduce AIGC traces.
npx skills add https://github.com/keji94/content-scanner --skill aigc-harness
Please help me install this Agent Skill. Skill: aigc-harness Source: https://github.com/keji94/content-scanner/tree/main/aigc-harness Command: npx skills add https://github.com/keji94/content-scanner --skill aigc-harness
AIGC Harness closes the loop on AI-generated content issues by automatically diagnosing AI traces, precipitating new detection or replacement rules, performing targeted rewrites, and validating convergence until the content meets pass thresholds. It reduces manual effort for teams that must lower AI-detectable characteristics while preserving meaning and correctness.
Run the harness on the file at content_path with the workspace_path containing rules and context to diagnose, precipitate rules if needed, perform targeted rewrites, and validate until convergence.
High-intent search queries and answers about installing and using this skill.
To reduce AIGC traces automatically, this Skill diagnoses AI flavor, precipitates replacement rules, performs targeted rewrites, and validates convergence until content passes detection thresholds.
The workflow for rewriting AI-generated text involves a four-phase loop: diagnosing AI traces, generating candidate rules, executing context-aware segment rewrites, and verifying convergence using a fix agent.
Yes, AIGC detection and rewrite requires integration with a content-scanner, domain-config settings, a fix agent, and a workspace containing context and replacement rules for processing.
You can manage and learn new rules by precipitating deterministic and semantic candidate rules, appending replacement pairs to the replacements table for ongoing coverage improvement during the rewrite process.
The best way to fix high AIGC scores is segmented, context-aware processing with adjacent context, applying configurable revision modes like rewrite, anti-detect, or polish per segment.
Automated AIGC rewrite needs multi-round fix loops to repeatedly validate content against detection thresholds, ensuring the revised text achieves convergence and passes the content-scanner checks.
vikisingh23Configure enterprise agents with industry, country, and regulatory presets.
ProfPowellStandardize Lucide-based icons with a local <x-icon> component.
9thLevelSoftwareParse and normalize semantic intent flags for Legion build and review workflows.
JimmyMa99Draft and polish diverse text types with specified tone alignment.
