restoreRestore a model's state from a snapshot previously returned by 'snapshot' (pass it back verbatim as 'snapshot'). Overwrites the model's current base state. The snapshot must be for the same model.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The model id. | |
| snapshot | Yes | A snapshot object as returned by the 'snapshot' tool. |
Changes observed during successful MCP inspections. Dates show when Glama detected each change.
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate destructiveHint=true and readOnlyHint=false. The description adds the concrete detail that it 'overwrites the model's current base state', which clarifies the scope of the destruction. It also discloses the same-model constraint. This goes beyond what annotations provide without contradicting them.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two succinct sentences with no wasted words. The primary action and the critical constraint (verbatim snapshot) are front-loaded, making it easy to scan and act on.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 2-parameter tool with no output schema and annotations covering safety, the description covers all essential aspects: what it does, the input format, the overwrite effect, and the same-model requirement. Nothing critical is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% for both parameters, but the description adds valuable semantics: it explains the 'snapshot' must be passed back verbatim and that it must be for the same model. These instructions go beyond the schema's simple 'object' type and improve correct usage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb 'restore' and resource 'model's state', and clearly distinguishes from siblings by referencing 'snapshot' and the overwriting behavior. The constraint 'must be for the same model' adds specificity, making the purpose unmistakable.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly ties usage to a prior 'snapshot' call and specifies the input must be passed verbatim. Does not list alternatives or when-not-to-use, but the relationship to snapshot is clear enough to route the agent appropriately.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools have clearly distinct purposes, but get_audit explicitly subsumes get_history and explain, and get_state with paths overlaps get_field, creating minor selection ambiguity. The detailed descriptions help, but an agent could still reach for the wrong getter.
Naming is overwhelmingly consistent: snake_case with verb_noun structure and coherent get_/create_/delete_ clusters. Minor deviations like bare verbs (mutate, explain, restore, snapshot) and eval instead of evaluate prevent a perfect score.
27 tools is above the comfortable range and feels heavy, especially with several overlapping audit/state getters that could be consolidated. That said, the domain is broad enough that the count is defensible, so it is heavy but not chaotic.
The tool set covers the full model lifecycle well: create, validate, test, mutate, evolve, read, delete, plus snapshot/restore, audit, blobs, views, library, and expression evaluation. Minor gaps like explicit export/import or separate view-management tools are workable around.