Skip to main content
Glama
Sign In

SKEIN

by net.skein
OverviewDiscussionsAdmin

skein_theme

Destructive

The caller's visual theme — colours, fonts, sizes, spacing. Call action="instructions" FIRST when the user asks for visual changes ("change the color", "another font", "make it darker", "like Alice's"): it returns the canonical fork → preview → confirm → persist workflow and the onboarding script, and improvising without it is the documented way to get this wrong. Built-in light/dark are read-only, so editing always starts with action="fork". Read-only actions: instructions, describe.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "action": "<action>"
    },
    "name": "skein_theme"
  }
}

Input Schema

NameRequiredDescriptionDefault
nameNoName for the new theme; unique among your own themes (another user's theme of the same name never collides), 1..100 chars. Required for fork.
actionYesWhat to do: - "instructions" → the canonical theme-editing guide (workflow, onboarding script, anti-patterns). Returns markdown, not JSON; cache it for the session. - "describe" → the token schema, the caller's active theme, every saved theme, and their recent Inspect-mode clicks. - "fork" → copy a theme (or a builtin) into a new editable one and activate it for the caller. - "preview_variables" → push an ephemeral preview to the caller's open tabs; nothing is saved. Applies your map verbatim, and warns (response `warning` + `roles[]`) about any key the current role dictionary does not declare — such a key paints nothing. - "set_variables" → merge variables into a saved theme and persist them. Refuses the WHOLE patch with `unknown_theme_role` (+ `roles[]`, with the successor of each retired name) if any key is not a current role: nothing is written, because a dead key in a saved row paints nothing and gets silently deleted by the frontend's next migration write-back. - "inspect_clear" → drop the caller's accumulated Inspect-mode selections.
sourceNoWhat to fork: "active" (the caller's current theme, falling back to their builtin), "builtin-light", "builtin-dark", or a theme UUID from describe. Required for fork.
themeIdNoUUID of the theme to write to, as returned by fork or describe. Required for set_variables.
variablesNoPartial map of CSS variable name → value. Supplied keys overwrite, absent keys are preserved (unlike the REST PATCH, which replaces the whole map). Required for set_variables and preview_variables. Every key must be a role of the CURRENT dictionary (action="describe", or the summary in action="instructions"): the pre-redesign names have no alias layer and resolve to nothing. set_variables refuses such a patch outright (`unknown_theme_role`); preview_variables applies it as sent and warns.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.3/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes far beyond the annotations (readOnlyHint=false, destructiveHint=true, idempotentHint=false): it discloses which built-ins are immutable, that preview_variables is ephemeral and unsaved, that set_variables merges rather than replaces (unlike REST PATCH), that a bad key refuses the whole patch atomically with unknown_theme_role, and that dead keys get silently deleted by frontend migration. That is exactly the mutation-risk context the annotations can only flag, not explain.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the primary rule (call instructions first) and grouped by read-only vs mutating actions; nearly every clause carries a constraint or failure mode. It is a dense single block rather than scannable structure, which costs a point on structure.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a six-action, no-output-schema tool this covers the decision path, immutability constraints, and error contracts well. It does not describe what describe/inspect_clear return, but the action enum in the schema already carries return shapes, keeping the gap minor.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so per-parameter meaning is already documented in the schema. The tool description adds only the ordering constraint (fork before editing) rather than extra semantics for source, themeId, or variables, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description names the resource concretely (visual theme: colours, fonts, sizes, spacing) and identifies the read-only actions, so an agent knows this is a theme-management surface. It is somewhat diffuse because the tool multiplexes six actions and the description leads with workflow guidance rather than a single verb.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit when-to-use trigger: "Call action=\"instructions\" FIRST when the user asks for visual changes" with real utterances ("change the color", "like Alice's") and an explicit warning that improvising is the documented way to get it wrong. It also states the when-not/alternative rule: built-in light/dark are read-only, so editing must start with action=\"fork\".

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All SKEIN toolsInspect
Try in Browser
Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources