Reduces a set of MCP tool definitions, grouped by server or flat, to the names-only index mcptoon puts in front of an agent — one line per server listing its tool names, no schemas — with the estimated token drop against the text you supplied. ...

Encodes one JSON document as JSON, TOON and Compact renderings and returns each form's character count, estimated token count and percentage saved against the JSON baseline — the way to measure mcptoon's output compression on a payload you actually have. ...

Parses TOON text into a JSON value and returns it, together with a canonical serialisation for comparing against the original — the way to prove an encode_toon round trip is lossless. Use it on text this server produced or on a stored TOON document; to produce such text use encode_toon, and to compare formats rather than parse use compare_formats. ...

Describes the running demo server itself: mcptoon version, the MCP protocol versions it accepts, the interpreter and platform versions behind it, and whether it carries any third-party runtime dependency. ...

Returns the submitted message text back unchanged, in one text content block, as a reachability probe for this server's stdio framing. Use it to confirm the MCP handshake and the transport work before testing anything else; it produces no analysis of its own, so for size comparisons call compare_formats and for a token figure call estimate_tokens. ...

Converts a JSON document into TOON, the line-oriented format mcptoon can serve in place of JSON, and returns the encoded text. Use it when you need the encoded form by itself; to see it next to the other formats with savings percentages use compare_formats, and to read such text back use decode_toon. ...

Counts the tokens in a text you supply using the estimator this package ships — character count divided by four, floored — so its numbers are directly comparable with the savings reported by build_tool_manifest and compare_formats. ...

Reports the AI agent configuration targets mcptoon's sync command knows how to write, each with the path it resolves to in this process's environment. ...

Returns this project's committed token measurements for 5, 50 and 255 MCP tool definitions encoded with tiktoken cl100k_base: raw JSON, TOON, SLIM and Compact token counts with each format's percentage saved. ...

Rewrites a single MCP tool definition (name, description, inputSchema) into mcptoon's slim form, where each parameter keeps only its name, its type and its required marker, and returns the slim definition with the character and estimated-token delta against what you sent. ...

Checks a proposed mcptoon server configuration before you save it and reports problems per entry: an entry with neither command nor url, one that sets both, args or env of the wrong JSON type, keys mcptoon ignores, and which transport mcptoon would choose. ...

List the upstream tools this gateway currently exposes, grouped by server. Returns names by default; that compact view is what an agent should read instead of pulling every full schema. Pass include_descriptions for a one-line summary per tool.

Report the MCP servers configured behind this gateway: transport, tool count and whether their tools loaded. Use it to see what is actually reachable before calling a tool, and to spot a server that configured but published nothing.

Self-check of this gateway process: version, config path, servers configured, tools indexed and uptime. Read this first when a tool you expect is missing; it distinguishes an empty config from a server that failed to load.