Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault

No arguments

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

CapabilityDetails
tools
{
  "listChanged": true
}
logging
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
checkA

Report the writing tells in text without changing a word. Use it to find where a draft carries the patterns the profile lists, and which ones. Each finding carries the rule that matched, the exact text, its line and column, and the swap the chop tool would make. It also returns a lint score from 0 to 100 measuring how densely those patterns appear. The score is not a reading on who wrote the text and must not be reported as one, and it compares only within one scoring version, which the reply names. Deterministic, free, and local: nothing is uploaded.

chopA

Rewrite text so it reads like a person wrote it instead of an AI. Use it to de-slop or humanize AI-generated text, remove AI tells from a draft, strip em-dashes, or hold writing to a house style. It cuts em-dash splices, semicolon habits, hedging and filler transitions, stock openers such as "In summary", the listicle reflex, and buzzwords such as "delve", "leverage", and "robust". The rules pass is deterministic, free, and runs on this machine with nothing uploaded, so the same input always gives the same output. Markdown code blocks and inline backtick spans are left alone. Returns the cleaned text, the tells it found, the tells it left for you to reword by hand, and the slop score before and after.

driftA

Measure a draft against the user's own writing and report where it stops sounding like them. Use it before handing over a draft you wrote for them, or to check whether text matches their voice. It compares sentence rhythm, punctuation habits, and register against the fingerprint the user measured from their own writing, and names each trait that landed outside their range, such as longer sentences or a heavier vocabulary. Drift is not slop: a trait outside the range can be good writing that is not theirs, so nothing is rewritten. It needs a fingerprint, which the user makes by running slop-chop voice fingerprint on their own writing. Deterministic, free, and local.

presetsA

List the built-in preset packs the chop and check tools accept. A preset overlays extra rules on the standard profile, such as cleaver for the aggressive swaps or plain for turning corporate phrasing into plain English.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

A4.1/5.0

Scored across 4 tools

Disambiguation5/5

Each tool targets a distinct function: check reports tells without editing, chop rewrites, drift compares against the user's own fingerprint, and presets lists overlays. The check/chop pair could superficially overlap but the descriptions sharply separate 'report' from 'rewrite'. No tool boundaries are ambiguous.

Naming Consistency4/5

All four names are single-word, lowercase, and stylistically uniform (check, chop, drift, presets), which is internally consistent even without a verb_noun pattern. The mix of verbs and a noun is a minor deviation from a predictable convention but remains readable and memorable.

Tool Count4/5

Four tools is on the thin side but each covers a genuinely distinct operation (lint, rewrite, voice-drift, presets) with no redundancy. It is slightly under-scoped rather than bloated.

Completeness3/5

The core analyze/rewrite loop is covered, but drift depends on a fingerprint the user makes by running 'slop-chop voice fingerprint', which is not exposed as a tool. That missing operation is a real dead end for agents, and there is no profile/preset-management surface beyond listing.

Maintenance

ActivityActive
ResponsivenessNo issues