Skip to main content
Glama

ZeroWidth Ledger

Record a decision, lesson, or observation in the Ledger

ledger_entries_create

Records a memory entry. Every entry is a TITLE (summary: one short plain sentence) over a BODY (rationale: the detail, markdown welcome) — never put the detail in the title. Three kinds: decision — a change being made now; PRE-REGISTRATION IS THE POINT, so prediction (what we expect) must be written NOW, before any evidence exists, and never backfilled to match an outcome. lesson — a distilled belief the team already holds (lesson text required, no prediction). observation — a durable fact worth remembering (no prediction): something already true, never something planned. Ideas, pitches, backlog items, and upcoming work are NOT entries — a dated piece of work that carries out a decision is a plan item (ledger_plan_add on that decision), and a running list you keep across runs belongs in a Napkin doc or sheet. When a user states something durable about their business in conversation, offer to capture it as a lesson or observation. When a user shares MEETING NOTES, propose the decisions you find with ledger_entries_draft so they keep or drop each one in Ledger; use this tool for an entry they've asked you to record. Link the Compass pages the entry touches so future consult-before-acting finds it. May return needs_confirmation — summarize the entry and wait for approval.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
kindNodecision = a change with a pre-registered expectation; lesson = a belief arriving already settled; observation = a fact that is already true (not a plan, idea, or pitch).decision
lessonNoWhat we learned — required for kind=lesson, optional for kind=observation, forbidden on decisions (their lesson is written at settlement).
summaryYesThe entry's TITLE: one short plain sentence naming what we did (decision) or the fact itself (lesson, observation). Plain text, no markdown, at most 280 characters — e.g. "Newsletter moves to a biweekly cadence". Everything longer goes in rationale.
rationaleNoThe entry's BODY: the detail under the title — why, context, lists, steps. Markdown is fine here and renders as formatted text.
workspaceNoWorkspace slug. Personal tokens with no default workspace MUST pass this; tokens with a default can override per call. Ignored for workspace API keys.
approvalIdNoApproval id from a prior needs_confirmation envelope.
predictionNoDecisions ONLY. What we expect: a list of `claims` plus one shared `deadline`, or `freeform` for room decisions nothing can settle. A claim either REACHES a value (comparator + target, e.g. CTA clicks >= 400) or HOLDS one (`hold`, e.g. newsletter reads no more than 5% below the 28 days before this). Name every number the change is expected to move AND every number it shouldn't cost — a decision that claims only what it hopes will rise gets to pick its own evidence. Set `watch: true` on a number worth following that the decision isn't committing to. When the change aims at part of an event metric ("new users in Germany"), narrow the claim with `slice` ({ country: ["DE"] }) using label values from ledger_metrics_get — a claim on an event metric settles on the total counted since landing (reach) or events per day (hold).
compassPageIdsNoWhere — Compass page ids this decision touches.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.5/5.0
Behavior4/5

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

Annotations declare readOnlyHint=false, destructiveHint=false, openWorldHint=false, so the write/safety profile is largely covered. The description adds genuinely non-annotation behavior: it may return `needs_confirmation`, in which case the agent should summarize and wait for approval, plus the pre-registration invariant (prediction written before evidence, never backfilled) that constrains correct use.

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 title/body model and the three kinds, and each subsequent sentence carries a rule or a routing decision rather than filler. It is dense and uses heavy caps for emphasis, which is slightly noisy, but nothing is padding.

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 an 8-parameter tool with nested prediction objects and no output schema, the description covers kinds, the pre-registration constraint, the approval flow, and the Compass linking purpose. Return-shape detail is the only meaningful omission, and the needs_confirmation envelope is at least flagged.

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, but the description adds framing the schema does not: the title/body split (summary as TITLE, rationale as BODY, never detail in the title) and the rule that a decision's lesson is written at settlement. The prediction block is well covered by its own schema description, so the added value is modest rather than transformative.

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 (records a memory entry in the Ledger) and immediately breaks the resource into its three kinds (decision/lesson/observation) with a one-line definition for each. It further distinguishes itself from siblings by name (ledger_plan_add, ledger_entries_draft), so an agent can route without opening another schema.

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 when-to-use routing: use ledger_entries_draft for meeting notes so the user keeps/drops each proposal, use ledger_plan_add for dated work that carries out a decision, use this tool for an entry the user asked to record, and put running lists in a Napkin doc. It also names what is NOT an entry (ideas, pitches, backlog items). This is about as complete as usage guidance gets.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources