aislopmeter
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| AISLOPMETER_CHROME | No | If neither Chrome nor Edge is found, set `AISLOPMETER_CHROME` to a browser path, or scan the project folder instead. | |
| AISLOPMETER_REPORT_DIR | No | Reports are written to `~/.aislopmeter/reports/`. Set `AISLOPMETER_REPORT_DIR` to change that. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| prompts | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| slop_scanA | Scan a website URL, a local project folder (Next.js, Vite/React, Astro, Vue, Svelte, plain HTML, Markdown content…) or a single .html file for AI slop: generic copy, cliché phrasing, template "vibe-coded" design, placeholder leftovers and missing trust signals. Returns a 0-100 slop score, five category scores and findings ranked by impact, each with evidence, a fix and (for folders) file:line locations to edit, plus a separate AI search readiness score (can ChatGPT, Claude, Perplexity and Gemini read and quote it: JSON-LD, robots.txt for AI crawlers, llms.txt, sitemap). If the project has a DESIGN.md, styles outside it are reported as design-drift. Re-scanning the same target automatically shows a before/after comparison. Prefer scanning the project folder when the user is working on the site's code; scan the URL for live sites. |
| slop_scan_htmlA | Scan raw HTML (for example a page the user pasted, or HTML you generated) for AI slop. Same report as slop_scan. Use slop_scan for URLs and folders. |
| slop_rescue_planA | Turn a scan into a prioritised, site-specific transformation plan: which facts to ask the owner for, quick clean-ups, copy rewrites, a design direction and trust fixes, each tied to findings, file:line locations and an estimated score gain. Follow the plan, then call slop_scan again to measure the change. |
| slop_compareA | Show the before/after difference between two scans: score change, resolved findings, remaining and new ones. |
| slop_reportA | Write the visual HTML report for a scan (a Report screen with the score and findings, and a calm What-to-do screen with a checklist, time estimates and the owner questions) and optionally open it in the user's browser. Use it at the end of a rescue so the user finds the result waiting when they return. Pass open: true only if the user agreed to have it opened. |
| slop_pageA | The full report for a single page of a site scan: its own score, categories and findings, each with steps and a done-when check. Use it to work through a site page by page. |
| slop_fixA | Concrete, step-by-step instructions for one finding of a scan: what to change, the exact file:line places, a before/after example and a "done when" check. Without finding_id it picks the most important open finding and lists the others. Use it to work through a scan one problem at a time. |
| slop_guideA | Read the reference guides behind AI Slop Meter: "criteria" (every check, its threshold and sources), "playbook" (the full rescue method), "copy" (rewriting rules with examples), "design" (how to leave the template look behind), "interview" (questions that get real facts from the owner). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| rescue | Scan a site or project, plan the transformation with the owner, apply it and re-measure until the target score. |
| fix | Scan, fix the single most important finding, re-scan and show what changed. Small, safe steps. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| aislopmeter-criteria | docs: docs/CRITERIA.md |
| aislopmeter-playbook | docs: playbook/RESCUE.md |
| aislopmeter-copy | docs: playbook/COPY.md |
| aislopmeter-design | docs: playbook/DESIGN.md |
| aislopmeter-interview | docs: playbook/INTERVIEW.md |
TDQS
Scored across 8 tools
Most tools have clearly distinct purposes, but slop_scan and slop_scan_html overlap in scanning HTML content, and slop_compare may be confused with the automatic before/after from re-scanning. The descriptions clarify these boundaries well enough to avoid misselection.
All tools use a consistent slop_ prefix and snake_case, making the set predictable. However, the parts of speech vary (verbs like scan vs nouns like guide) and slop_scan_html adds an _html suffix that slightly breaks the single-word pattern.
Eight tools is well within the ideal range for a domain-specific server. Each tool earns its place by covering a distinct step in the scan-fix-report-guide workflow, from initial scanning to final reporting.
The surface covers scanning multiple input types (URL, folder, raw HTML), per-finding fixes, per-page reports, comparison, rescue planning, and reference guides. This provides a complete lifecycle for identifying and remediating AI slop without obvious gaps.