Skip to main content
Glama

Savepoint before an instruction

figma_savepoint
Idempotent

Mark the point before an instruction, so figma_rollback can take back everything done after it in one call. Make one at the start of every instruction a person gives: it writes nothing to the document and costs nothing.

Instructions

Mark the point before an instruction, so figma_rollback can take back everything done after it in one call. Make one at the start of every instruction a person gives: it writes nothing to the document and costs nothing. Result { savepointId, label, at, pageId, journalSeq }; figma_get_changes { since: savepointId } lists what was done since. Savepoints live in this server process: a restart forgets them, and a rollback forgets the ones made after its own. Editors: Design; unverified: FigJam, Slides.

Input Schema

NameRequiredDescriptionDefault
forceNoOverwrite on purpose. A write to a node someone changed after you last read it is refused (STALE); read it again and resend instead. Pass true only when the person said to overwrite their edit.
labelYesThe instruction in a few words, e.g. "Hero copy from the brief".
sessionNoOptional sessionId or file name, to pick a Figma file when more than one is connected.
idempotencyKeyNoCaller-owned retry key, ≤128 chars. Re-calling with the same key never executes twice: it returns the existing result, or waits for the op still in flight. Supply your own for anything large — if the MCP client times out (rather than the bridge), only a key you chose can be used to retry safely.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Addedv0.3.0

TDQS

A4.6/5.0
Behavior5/5

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

Discloses state semantics annotations cannot express: 'Savepoints live in this server process: a restart forgets them, and a rollback forgets the ones made after its own.' It also declares the result shape ({ savepointId, label, at, pageId, journalSeq }) and scopes editor support (Design verified; FigJam and Slides unverified), which is exactly the caution an agent needs. Nothing here contradicts readOnlyHint=false — the tool writes no document content but does create server-side state.

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?

Purpose and the rollback linkage are front-loaded in the first sentence, and every subsequent sentence adds distinct value (cost, return shape, lifecycle, editor coverage). It is dense — four sentences carry purpose, guidance, output, caveats and platform scope — but none is filler.

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?

With no output schema, the description supplies the return fields itself, defines the lifecycle limits (restart and rollback invalidation), and flags editor coverage. For a stateful, non-destructive tool with one required parameter, an agent has everything needed to decide to call it at the right moment and reuse the result correctly.

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

Parameters3/5

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

Schema description coverage is 100%, so force, label, session and idempotencyKey are already well documented; baseline 3 applies. The description adds only indirect value by showing savepointId is the token later passed to figma_get_changes' `since`, which makes the returned ID's role concrete but does not explain any parameter itself.

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 ('Mark the point before an instruction') and immediately ties it to the sibling that consumes it ('so figma_rollback can take back everything done after it'). An agent can distinguish this from figma_rollback, figma_get_changes and figma_undo_step without opening any 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, actionable guidance: 'Make one at the start of every instruction a person gives', plus the cost argument ('writes nothing... costs nothing') that removes hesitation. It also names the downstream alternatives and their relationship — figma_rollback for undo and figma_get_changes { since: savepointId } for inspection.

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

Deploy Server

Other Tools