Taseli

Remember for this business

remember

Idempotent

Save one thing to Taseli's memory of this business so every later decision follows it: the plan, the daily review, new ads and campaigns, every answer. Holds one lasting rule or preference ('always…', 'never…', 'I prefer…', 'from now on…', 'don't touch…'), a fact about the business that isn't in the business card (a new branch, a price, a season, a product they stopped selling), or a decision the owner made against advice. One note per call. Not for questions, passing remarks, or numbers that change daily. A setting Taseli already has (budget, target, goal, brand words, protected campaigns) is changed through a proposal the owner approves in Taseli; a note here holds only the reason behind it, if one was given. Changes nothing in any ad account.

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": {
      "statement": "<statement>",
      "kind": "<kind>",
      "said_by_owner": "<said_by_owner>"
    },
    "name": "remember"
  }
}

Input Schema

Name	Required	Description
`say`	No	The same note, short, in the language of the conversation, as the owner would read it (shown to them and in Settings › What Taseli knows).
`kind`	Yes	fact: about the business. preference: how the owner wants things done. decision: a choice the owner made, also against advice.
`quote`	No	The owner's own words, as they wrote them (any language). Required when said_by_owner is true.
`topic`	No	A short stable key, lowercase with dots, e.g. 'pref.shabbat', 'offer.price.cleaning', 'area.branches'. A later note with the same topic replaces this one, so reuse the topic when the owner changes their mind.
`account`	No	An ad account's name, when it is about one account only; omit when it is about the whole business.
`cost_note`	No	With against_advice, in the chat: the expected cost in one short sentence in the owner's language. It is shown once as a quiet note under the saved line.
`statement`	Yes	One plain English line that stands on its own, e.g. 'Never run ads from Friday 16:00 to Saturday night (Shabbat).' or 'Haifa branch opened in October 2026; same services as Tel Aviv.'
`said_by_owner`	Yes	true only when the owner said it outright in this conversation; false when it was inferred rather than said.
`against_advice`	No	Only for a decision against best practice: in one English line, what you advised and the expected cost, e.g. 'Advised keeping the brand campaign (about 30% of sales); owner turned it off.'
`for_all_clients`	No	Only for an agency's own team, for a preference: true when they state how the agency works for all its clients ('we always…', 'in all our accounts…', 'never for any client…'): naming, structure, bidding, budget and approval habits, creative tone, report format, never-do's (topic: agency.naming, agency.structure, agency.bidding, agency.budget, agency.approval, agency.creative, agency.report or agency.never). Saved in the agency's memory and followed in every client's account unless a client's own rule says otherwise. Not used for a rule about one business.

Output Schema

Name	Required	Description
`note`	No	A plain note instead of data (nothing connected yet, not available here).
`text`	No	The answer in plain words.
`truncated`	No
`truncatedNote`	No

Schema Changelog

Changes observed during successful MCP inspections.

First observedNov 3, 2026

TDQS

A4.6/5.0

Behavior4/5

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

Annotations declare a non-read-only write with no destructiveness, and the description usefully bounds that write ('Changes nothing in any ad account') plus a cardinality rule ('One note per call'). It does not cover replacement/overwrite semantics in the description text (that lives in the topic parameter) or any permission/rate context, so it adds real value but not a full behavioral picture.

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-loads the verb and the persistence guarantee, then layers inclusion, exclusion, and the boundary case, so an agent can stop reading early and still act. The third sentence is long and parenthetical-dense, and the 'Not for…' clause would read better as its own line, but nothing is genuinely redundant.

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

Completeness5/5

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

For a 10-parameter write tool with an output schema and full annotation coverage, the description supplies what the structured fields cannot: what belongs in memory, what does not, one-note-per-call discipline, and the fact that no ad account is touched. Return values are covered by the output schema, so no gap remains that would cause a mis-call.

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 100%, so the baseline is 3; the description earns an extra point by giving concrete instantiations of the 'kind' enum ('always…', 'never…', 'I prefer…' versus a fact like a new branch or a discontinued product) and by framing the against-advice decision case that maps to against_advice. It still says nothing about say, quote, topic, account, or for_all_clients, leaving those entirely to the schema.

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

Purpose5/5

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

States a specific verb and resource ('Save one thing to Taseli's memory of this business') and immediately scopes the effect ('so every later decision follows it'). It is clearly distinguishable from the read-side sibling get_memory and from record_idea, since it enumerates exactly what qualifies as a memory item.

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?

Gives explicit inclusion criteria (lasting rules/preferences, non-card business facts, decisions made against advice) and explicit exclusions ('not for questions, passing remarks, or numbers that change daily'). It also routes a competing case elsewhere: a setting Taseli already owns is changed via an owner-approved proposal, and only the reason behind it belongs here.

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

All Taseli tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?