Skip to main content
Glama

Save this chat to RAMen

save_thread
Idempotent

Save the current chat's state to continue in a fresh chat or another AI app without re-reading the full thread, cutting token use and keeping key details.

Instructions

Save where this conversation stands, so it can go on in a fresh chat or in another AI app (Claude, ChatGPT or any app with RAMen Memory) without re-reading this whole chat: every reply in a long chat re-reads all of it, which uses up tokens and usage limits. Use it when the user asks to save the chat or to continue it in a new chat or elsewhere, or when the chat has grown long. Be complete and specific: every name, number, date, time, price and reason, every change (with what it was before), and the latest version of anything you wrote together, word for word. Saving again under the same name adds to the thread and updates its values.

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": {
      "thread": "<thread>",
      "goal": "<goal>",
      "now": "<now>",
      "facts": "<facts>",
      "decisions": "<decisions>",
      "open": "<open>",
      "drafts": "<drafts>"
    },
    "name": "save_thread"
  }
}

Input Schema

NameRequiredDescriptionDefault
nowYesvalues that can change (dates, counts, budgets, choices, status): the current value, and the one before if it changed
doneNoopen items finished since the last save (their words)
goalYeswhat this work is for, in one sentence
openYesthe open questions, to-dos and deadlines now (this list replaces the saved one)
factsYeseverything else that holds: names, places, times, prices, numbers, people's roles, constraints, preferences; one specific, self-contained sentence each
draftsYesevery piece written together (emails, messages, lists, plans, wording), each in its latest version, word for word; do not leave drafts out to save space: they are what a new chat cannot rebuild
threadYesa short name for this work, for example 'wedding planning'
projectNothe project the user puts this work in, if they name one, for example 'wedding'
decisionsYeseach decision, why, and what was turned down
checkpointNotrue for a save you make on your own about every 15 messages in a long chat (the user's "Save long chats as they go" setting decides; RAMen answers 'Not saved' when it is off)

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.0.1

TDQS

A4/5.0
Behavior4/5

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

Annotations only declare read/write and safety hints; the description goes further by explaining why saving matters (token and usage-limit cost of long chats), what a same-name re-save does to existing content, and how the checkpoint flag interacts with the user's "Save long chats as they go" setting, including the "Not saved" response when it is off. Minor tension worth noting: the append/merge behavior sits uneasily beside idempotentHint=true, though the honest reading is an upsert rather than a contradiction.

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 action and its benefit before the trigger conditions and content rules, and every clause carries information. It is a single dense run-on paragraph, though, and the completeness instructions run long relative to the trigger guidance.

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 10-parameter, 7-required mutation tool with no output schema, the definition covers when to call, what must be captured, and how re-saves merge, which is most of what an agent needs. The gaps are small but real: nothing about what a successful save returns, and no mention of the `project` or `done` fields despite their role in organizing the thread.

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

Parameters4/5

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

Schema coverage is already 100%, so baseline is 3, and the description earns above that by adding substantive content rules: include every name, number, date, price and reason, every change with its prior value, and keep drafts word for word because "they are what a new chat cannot rebuild." It does not discuss the `project` or `done` parameters, leaving those entirely to the schema.

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?

States a specific verb and resource (save this conversation's state) plus the payoff: continuing in a fresh chat or another AI app without re-reading the whole chat. It clearly separates itself from a pure recall/resume tool by framing itself as the write side, though it never names the sibling tools it pairs with.

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

Usage Guidelines4/5

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

Gives concrete triggers: the user asks to save the chat, asks to continue elsewhere, or the chat has grown long. It also covers the re-save case ("saving again under the same name adds to the thread and updates its values"), which is a real ambiguity resolved. No explicit when-not guidance or naming of recall_thread/resume_thread as the retrieval counterpart.

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