EMET

emet_session_close

Verify a session's work is written to the store before closing: checks receipts, reports missing items, and blocks or escalates incomplete closes; it writes nothing itself.

Instructions

Call this when the session is ending - the user says goodbye, asks you to wrap up, mentions restarting or opening a new session, or you are about to run out of room. An acknowledgement of the turn just completed - thanks, great, perfect - is not a session end; if it is ambiguous, ask rather than close. Checks whether this session's work has actually been written to the store and reports exactly what is missing. It does not write anything itself. RECEIPTS ARE REQUIRED: pass receipts naming what you wrote and read back for each close step - working documents, the episodic session record id, the board and its version, the transcript part ids and highest exchange index as query_transcripts returned them, and the handoff version (written last). Each receipt is compared with the store; omitted or mismatched receipts, a handoff that was not the last write, or a transcript no layer entry cites (derived_from) return state 'blocked'. Each missing line carries a scope: touch only what it names. THREE blocked closes on one session return state 'escalate': stop retrying, say the speak lines, and hand the user the list - the counter does not reset on a changed receipt, only with a new session. Say every line of the returned speak to the user, in order - a wrap that leaves it out is a failed close. Never tell the user their session is saved unless this returns state 'complete'.

Input Schema

TableJSON Schema

Name	Required	Description
`source`	No	Your host's source tag: the registered tag this session was written under. The close is never refused for it; an unregistered tag, one that did not write this session, or none (with no server default) gets a warning that the close is not attributed.
`exchange`	No	The exchange (turn) this write belongs to, counting from 1. A write for exchange n while exchange n-1 was never appended is refused by the session graph until n-1 is appended.
`receipts`	No	Required for state 'complete'. One entry per close step, taken from what the tools returned this session - not from recall.
`session_id`	No	The session_id used (or to be used) for this session's transcript, e.g. '2026-09-03_short-description'. Without it the transcript cannot be checked; receipts.transcript_session_id is used when this is omitted.
`thread_docs`	No	Additional document ids this session should have updated, e.g. ['threads/some-project.md'].
`session_start`	No	Unix seconds when the session began. Defaults to 12 hours ago, which makes the answer approximate.

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedv0.2.2Oct 5, 2026

changedInput schema / properties / exchange / description

Previous value: -"The exchange (turn) this write belongs to, counting from 1. A write for exchange n while exchange n-1 was never appended is flagged by the session graph."New value: +"The exchange (turn) this write belongs to, counting from 1. A write for exchange n while exchange n-1 was never appended is refused by the session graph until n-1 is appended."

addedInput schema / properties / source

Added value: +{
+  "description": "Your host's source tag: the registered tag this session was written under. The close is never refused for it; an unregistered tag, one that did not write this session, or none (with no server default) gets a warning that the close is not attributed.",
+  "type": "string"
+}

First observedv0.1.0Oct 3, 2026

TDQS

A4.6/5.0

Behavior4/5

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

It discloses behavior far beyond the annotations: the receipt-comparison protocol, that omitted/mismatched receipts, a non-last handoff, or an uncited transcript yield state 'blocked'; that three blocked closes yield 'escalate' and that the counter persists across receipt changes; and that every 'speak' line must be relayed. One mild tension: readOnlyHint=false implies side effects while the text says it 'does not write anything itself', though the persistent blocked-close counter plausibly accounts for the non-idempotent annotation.

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?

The text is long but dense, and the most consequential rules are front-loaded (when to call, then what it does, then RECEIPTS ARE REQUIRED) with emphasis markers for the blocked/escalate branches. A few clauses, such as the repeated warnings about relaying speak lines, could be tightened, but almost every sentence carries an operative rule.

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?

Despite a complex 6-parameter schema with a nested receipts object and no output schema, the description covers the return states ('complete', 'blocked', 'escalate'), the per-line 'scope' field, and the required 'speak' output, so an agent knows both how to call it and how to act on the result. Nothing essential is left unspecified.

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 real semantics the schema cannot: receipts are required for state 'complete', each receipt is compared against the store, and each missing line carries a 'scope' telling the agent to touch only what it names. It also stresses receipts must come from tool returns, not recall, which is meaningful usage guidance for the nested object.

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?

The description states a precise verb+resource: it 'Checks whether this session's work has actually been written to the store and reports exactly what is missing' and explicitly disclaims writing ('It does not write anything itself'). This distinguishes it cleanly from sibling write tools like save_to_layer, write_doc, and emet_transcript_append, and from the sibling emet_status.

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?

It gives explicit triggering conditions ('user says goodbye, asks you to wrap up, mentions restarting or opening a new session, or you are about to run out of room') and an explicit exclusion ('An acknowledgement of the turn just completed - thanks, great, perfect - is not a session end'), plus a tie-breaker ('if it is ambiguous, ask rather than close'). This is exactly the when/when-not guidance the dimension rewards.

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

emet_session_close

Instructions

Input Schema

Schema Changelog

TDQS

Other Tools