A red-pen pass leads both invocation branches
The model-facing description now covers drafting new prose and revising supplied text without repeating the old metadata trigger.
Stop Slop Skill ยท Red-pen editing
Keep the writer's meaning. Cut the predictable phrasing.
The optimized Stop Slop workflow locks facts and voice, opens only the references a draft needs, resolves every flagged span, and repeats the weakest review dimensions until the existing 35/50 threshold is met.
Stop Slop is Hardik Pandya's red-pen editing Skill for preserving source facts, voice, required terms, and uncertainty while removing predictable AI phrasing and formulaic rhythm.
The model-facing description now covers drafting new prose and revising supplied text without repeating the old metadata trigger.
Facts, claims, names, numbers, dates, quotations, audience, voice, links, formatting, length limits, and uncertainty are recorded before changing words.
Phrase, structure, and example references load only when a flagged span needs them; every flag must resolve or carry a reason before the weakest dimensions repeat.
The optimized Skill keeps the upstream scoring bar instead of inventing a new quality scale or claiming authorship detection.
Keeps the writer's facts, voice, required terminology, links, formatting, and uncertainty in view before editing.
Loads phrase, structure, and example references only when the draft needs their guidance.
Changes the flagged span while preserving precise technical language, quotations, and deliberate voice choices.
Scores directness, rhythm, trust, authenticity, and density, then repeats the weakest dimensions as needed.
Keep the supplied offer, facts, and voice while replacing formulaic copy.
Protect exact terms, paths, and constraints while removing vague or repetitive scaffolding.
Build a source lock from the brief before drafting and reviewing the result.
Verified source-diff evidence
This release proves canonical source lineage, package inventory, the optimized true diff, and an immutable artifact. It does not claim runtime outcomes or visual comparisons.
.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/original-inventory.json#/entries?sha256=6441805b9f3d7e703465ab3a31eb63b31687be1f6f772bd27c467d8304b07602.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/optimized-inventory.json#/entries?sha256=17c60eba11e4465a653c4fd33a8a7e204a3061d822d17113c27f597810a807a8.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json#/evidenceUnits/0?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json#/evidenceUnits/1?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json#/evidenceUnits/2?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json#/evidenceUnits/3?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff.scratch/kan-300/production-release/source-evidence/candidates/stop-slop/change-set.json?sha256=c17d394eba86778eb306fa1df00d1060d79528fab21df1d15c94081d9477f1ff# Stop Slop This edition derives from Hardik Pandya's stop-slop package at 8da1f030185bdfe8471220585162991eaeb970e9. ## Four-gate workflow Lock the source, draft or diagnose, rewrite against the lock, and audit the final prose. Load the phrase, structure, and example references only when their flags apply. Repeat the weakest dimensions until the existing 35/50 score is met. The seven-file repository package, README, changelog, references, and MIT license remain intact. This release edition intentionally has no real prose comparison.
Use these phrases to activate this skill in your AI coding assistant: