zero-slop 2.12.2 → 2.12.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (6) hide show
  1. package/README.md +28 -62
  2. package/SKILL.md +1 -1
  3. package/docs/cli.md +1 -1
  4. package/gemini-extension.json +1 -1
  5. package/package.json +1 -1
  6. package/plugin.json +1 -1
package/README.md CHANGED
@@ -1,40 +1,19 @@
1
- <p align="center">
2
- <a href="https://zero-slop.ai">
3
- <picture>
4
- <source media="(prefers-color-scheme: dark)" srcset="assets/logo/zero-slop-logo-reversed.svg">
5
- <img src="assets/logo/zero-slop-logo-primary.svg" width="360" alt="Zero Slop">
6
- </picture>
7
- </a>
8
- </p>
9
-
10
- <p align="center"><strong>Edit AI-assisted writing without losing the details.</strong></p>
11
-
12
- <p align="center">
13
- Zero Slop is an open-source skill that finds stock phrases, repetition, vague claims, and mechanical rhythm.<br>
14
- Your AI assistant edits the draft. Local tools check its names, numbers, links, quotations, code, tables, and paths against the original.
15
- </p>
16
-
17
- <p align="center">
18
- <a href="https://zero-slop.ai/try/"><strong>Try it in your browser</strong></a>
19
- ·
20
- <a href="#quick-start">Install the skill</a>
21
- ·
22
- <a href="#evidence-and-limits">See the evidence</a>
23
- ·
24
- <a href="https://github.com/manavmishra/ZeroSlop/releases/latest">Latest release</a>
25
- ·
26
- <a href="https://zero-slop.ai/brand/">Brand assets</a>
27
- </p>
28
-
29
- <p align="center">
30
- <a href="https://github.com/manavmishra/ZeroSlop/actions/workflows/validate.yml"><img alt="Validate" src="https://github.com/manavmishra/ZeroSlop/actions/workflows/validate.yml/badge.svg"></a>
31
- <img alt="Version 2.12.2" src="https://img.shields.io/badge/version-2.12.2-72528F?color=C15732">
32
- <a href="https://www.npmjs.com/package/zero-slop"><img alt="npm version" src="https://img.shields.io/npm/v/zero-slop?color=C15732"></a>
33
- <a href="https://www.npmjs.com/package/zero-slop"><img alt="npm downloads" src="https://img.shields.io/npm/dm/zero-slop?color=17634F"></a>
34
- <a href="https://github.com/manavmishra/ZeroSlop/stargazers"><img alt="GitHub stars" src="https://img.shields.io/github/stars/manavmishra/ZeroSlop?style=flat&color=C15732"></a>
35
- <a href="LICENSE"><img alt="MIT license" src="https://img.shields.io/badge/license-MIT-141412"></a>
36
- <a href="https://hol.org/registry/plugins/manav-mishra%2Fzero-slop"><img alt="Listed in the HOL plugin registry" src="https://img.shields.io/badge/HOL%20registry-listed-2C6E8F"></a>
37
- </p>
1
+ <picture>
2
+ <source media="(prefers-color-scheme: dark)" srcset="assets/logo/zero-slop-logo-reversed.svg">
3
+ <img src="assets/logo/zero-slop-logo-primary.svg" width="180" alt="Zero Slop">
4
+ </picture>
5
+
6
+ # Zero Slop
7
+
8
+ Find the canned phrases in an AI-assisted draft, then edit them without losing what you meant to say.
9
+
10
+ Zero Slop is a free, open-source agent skill. Your AI assistant reads and edits the writing; local tools score the result and check names, numbers, links, quotations, code, tables, and paths against the original. [Try it in your browser](https://zero-slop.ai/try/) or [install the skill](#quick-start).
11
+
12
+ <img alt="Version 2.12.4" src="https://img.shields.io/badge/version-2.12.4-72528F?color=C15732">
13
+
14
+ ## Problem
15
+
16
+ AI can give you a grammatically clean draft that still sounds like everybody else's: the same opening, the same tidy contrasts, the same closing thought. Asking for a rewrite can make things worse if the assistant quietly changes a number or smooths away your point. Zero Slop gives it an editing brief and checks the details afterward. A writing score helps you spot familiar patterns, but it cannot tell who wrote the text.
38
17
 
39
18
  <a href="assets/zero-slop-demo.mp4?v=dark-shell-restored-20260906">
40
19
  <picture>
@@ -50,7 +29,7 @@ Here is a short launch draft:
50
29
 
51
30
  > We're thrilled to announce that our team has leveraged cutting-edge machine learning to deliver a seamless onboarding experience, reducing setup time by 40%.
52
31
 
53
- The local scorer rates it 99.3/100. Lower is better. It flags “We're thrilled
32
+ The local scorer gives it 99.3/100. Lower is better. It flags “We're thrilled
54
33
  to,” “leveraged,” “cutting-edge,” and “seamless.”
55
34
 
56
35
  An edit that keeps the stated result:
@@ -62,7 +41,7 @@ Writing score: 9.5/100 [clear]
62
41
  Flagged phrases : 0 across 10 words
63
42
  ```
64
43
 
65
- The result still claims a 40% reduction in setup time. Scores describe writing patterns; they cannot identify the author. See four complete pairs in [`examples/`](examples/).
44
+ The result still claims a 40% reduction in setup time. See four complete pairs in [`examples/`](examples/).
66
45
 
67
46
  Use it on launch posts, changelogs, emails, or research summaries. You can score drafts without editing.
68
47
 
@@ -78,18 +57,11 @@ npx skills add manavmishra/ZeroSlop --global
78
57
  /zero-slop (your writing)
79
58
  ```
80
59
 
81
- To inspect without rewriting, use `/zero-slop inspect (your writing)`. Or [try the browser editor](https://zero-slop.ai/try/).
60
+ To find problems without rewriting, use `/zero-slop inspect (your writing)`. You can also [try the browser editor](https://zero-slop.ai/try/) before installing anything.
82
61
 
83
- ### Other ways to use it
62
+ ### Other ways to install
84
63
 
85
- | Environment | Fastest route |
86
- |---|---|
87
- | Claude Code, Codex, Cursor, OpenCode, Warp, Zed | `npx skills add manavmishra/ZeroSlop --global` |
88
- | Gemini CLI | `gemini extensions install https://github.com/manavmishra/ZeroSlop --auto-update` |
89
- | Claude Code plugin | `/plugin marketplace add manavmishra/ZeroSlop`, then `/plugin install zero-slop@zero-slop` |
90
- | Any assistant with file uploads | Download the [single-file bundle](https://github.com/manavmishra/ZeroSlop/releases/latest/download/zero-slop-single-file.md) |
91
- | Claude.ai | Upload the [latest skill ZIP](https://github.com/manavmishra/ZeroSlop/releases/latest/download/zero-slop.zip) |
92
- | ChatGPT, Claude, Grok, Gemini, Cursor, or another MCP client | Connect the optional [hosted MCP server](mcp/README.md) |
64
+ The `npx skills add` command works with Claude Code and Codex; [other compatible clients](DISTRIBUTION.md) have install notes. Gemini CLI users can run `gemini extensions install https://github.com/manavmishra/ZeroSlop --auto-update`. For Claude.ai, upload the [skill ZIP](https://github.com/manavmishra/ZeroSlop/releases/latest/download/zero-slop.zip); for file-upload assistants, use the [single-file bundle](https://github.com/manavmishra/ZeroSlop/releases/latest/download/zero-slop-single-file.md).
93
65
 
94
66
  For a local score without a model call:
95
67
 
@@ -103,7 +75,7 @@ From a cloned checkout, check a folder against the review threshold of 25:
103
75
  python3 scripts/slopscore.py --batch drafts/ --gate 25
104
76
  ```
105
77
 
106
- Installed checks run locally; skill editing follows your assistant's privacy settings. The [hosted MCP service](mcp/README.md) and browser editor process drafts remotely with Workers AI.
78
+ Installed checks run locally; editing through the skill follows your assistant's privacy settings. The [hosted MCP service](mcp/README.md) and browser editor process drafts remotely with Workers AI.
107
79
 
108
80
  For an MCP client, use this endpoint:
109
81
 
@@ -111,14 +83,14 @@ For an MCP client, use this endpoint:
111
83
  https://mcp.zero-slop.ai/mcp
112
84
  ```
113
85
 
114
- [Connection options and listing status](DISTRIBUTION.md).
86
+ The MCP connection options are in [`mcp/README.md`](mcp/README.md).
115
87
 
116
88
  ## Edit from the command line
117
89
 
118
90
  The CLI sends a file to the hosted editor without changing the file on disk:
119
91
 
120
92
  ```sh
121
- npx --yes zero-slop@2.12.2 deslop draft.md --genre professional
93
+ npx --yes zero-slop@2.12.4 deslop draft.md --genre professional
122
94
  ```
123
95
 
124
96
  Use `-` for stdin and `--json` for structured output. `--require-approved` prints the result but exits nonzero when review is needed. Requires Node.js 22+; offline `score` also needs Python 3. [CLI options and privacy](docs/cli.md).
@@ -131,13 +103,11 @@ curl --fail-with-body --max-time 75 https://mcp.zero-slop.ai/v1/deslop \
131
103
  --data '{"text":"Maya owns the pricing review.","genre":"professional"}'
132
104
  ```
133
105
 
134
- REST returns the same result as MCP. Check `status` before using an edit. Shared free capacity accepts up to 20,000 Unicode code points after trimming. Hosted CLI and REST editing process drafts remotely without storing them.
135
-
136
- [API reference](docs/rest-api.md) · [OpenAPI contract](https://mcp.zero-slop.ai/openapi.json)
106
+ REST returns the same result as MCP. Check `status` before using an edit. Shared free capacity accepts up to 20,000 Unicode code points after trimming. [API reference](docs/rest-api.md) · [OpenAPI contract](https://mcp.zero-slop.ai/openapi.json)
137
107
 
138
108
  ## What it does
139
109
 
140
- The installed skill ships no model. Your AI assistant, whether Claude, GPT, or another compatible model, reads and edits the draft. The skill supplies the workflow and local tools: a 0–100 writing score, source-detail checks, and a final comparison with the original. The eight stages divide the work.
110
+ The installed skill ships no model. Your AI assistant, whether Claude, GPT, or another compatible model, reads and edits the draft. The skill supplies the workflow and local tools: a 0 to 100 writing score, source-detail checks, and a final comparison with the original.
141
111
 
142
112
  Use the score to locate passages worth reviewing. Optional private learning records reason-labelled corrections you provide; it does not learn a complete writing style.
143
113
 
@@ -179,9 +149,7 @@ If repair fails, the editor returns the safest edit with a warning and stops the
179
149
 
180
150
  ### Optional reader review
181
151
 
182
- Ask: **“Review this for backend engineers. Where would they stop reading? Don't rewrite it.”** An optional skim, passage reactions, and notes-only recall return comments. Simulated reactions cannot establish human behavior. Sequential review needs isolated contexts; see the [privacy and reporting limits](references/reader-review.md). Hosted calls are unchanged.
183
-
184
- Inspired by [First Reader](bench/first-reader/), which only reviews. Zero Slop also rewrites and checks source details. Neither has established human-reader accuracy; First Reader has no invented rewrite score.
152
+ You can ask the skill where a particular audience might stop reading without requesting an edit. The [reader-review guide](references/reader-review.md) explains the limits: simulated reactions are not evidence of what real readers prefer. This optional mode adds no hosted model calls.
185
153
 
186
154
  ## Evidence and limits
187
155
 
@@ -214,9 +182,7 @@ Zero Slop's outputs came from v2.5.9; newer versions only rescore them. The curr
214
182
 
215
183
  On 75 labelled antithesis pairs, the reading pass reached 91.2% recall across the full set, 100% recall on shapes in reach, and 100% precision. We constructed and labelled them; this is a regression test, not field accuracy.
216
184
 
217
- Local timings exclude AI editing. On one Apple silicon Mac, the scorer processed 1,000 documents in a median 1.9929 seconds (501.8 per second), with five runs ranging from 1.9614 to 2.0945 seconds. A 15,201-word document took a median 0.3223 seconds; the slowest stress case took 2.2932 seconds; learning from an 8,000-word edit took 0.1592 seconds. [Machine and results](bench/performance-results.json).
218
-
219
- Across 12 interleaved runs against 2.7.7, median throughput was 0.26% lower, within the 5% regression limit. The two-way replay used Zero Slop v2.6.0. These local timings are not a service-level guarantee.
185
+ Local timings exclude AI editing. On one Apple silicon Mac, the scorer checked 1,000 documents in a median 1.9929 seconds (501.8 per second). Across 12 interleaved runs against 2.7.7, it had 0.26% lower median throughput, within the 5% regression limit. The two-way replay used Zero Slop v2.6.0. [Machine and results](bench/performance-results.json). These timings are not a service-level guarantee.
220
186
 
221
187
  The [RAID+ audit](bench/raid-plus-corpus/README.md) asks a different question: how often does the scorer flag writing from different models? The pinned sample contains 7,627 usable generations:
222
188
 
package/SKILL.md CHANGED
@@ -2,7 +2,7 @@
2
2
  name: zero-slop
3
3
  license: MIT
4
4
  metadata:
5
- version: "2.12.2"
5
+ version: "2.12.4"
6
6
  author: manavmishra
7
7
  description: Edit drafts into natural prose, inspect AI-sounding patterns, or review how a specified audience might respond passage by passage. Zero Slop runs inside the user's existing AI assistant with local tools that protect source details. Use for humanizing or de-slopping writing, polishing outward-facing prose, social drafts, final editorial checks, or an explicit simulated reader review. Preserve facts, voice and format; reader simulations are hypotheses, not human feedback.
8
8
  ---
package/docs/cli.md CHANGED
@@ -3,7 +3,7 @@
3
3
  Use `deslop` for hosted editing with MCP parity. Use `score` for offline checks.
4
4
 
5
5
  ```sh
6
- npm install --global zero-slop@2.12.2
6
+ npm install --global zero-slop@2.12.4
7
7
  zero-slop deslop draft.md --genre professional
8
8
  zero-slop deslop - --genre email < draft.txt
9
9
  zero-slop deslop draft.md --json --require-approved
package/gemini-extension.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "zero-slop",
3
- "version": "2.12.2",
3
+ "version": "2.12.4",
4
4
  "description": "Find and fix canned AI writing while preserving the facts, voice, and format.",
5
5
  "mcpServers": {
6
6
  "zero-slop": {
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "zero-slop",
3
- "version": "2.12.2",
3
+ "version": "2.12.4",
4
4
  "description": "An editing skill, offline writing checks, and a CLI for the hosted Zero Slop MCP pipeline. Find stock AI phrasing, rewrite, and check source details.",
5
5
  "bin": {
6
6
  "zero-slop": "bin/zero-slop.mjs"
package/plugin.json CHANGED
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
3
3
  "name": "zero-slop",
4
- "version": "2.12.2",
4
+ "version": "2.12.4",
5
5
  "description": "Find and fix canned AI writing while preserving the facts, voice, and format.",
6
6
  "author": {
7
7
  "name": "Manav Mishra",