`structured` and `toStructuredOutput` may expose a smaller validated result than the complete execution output. Dynamic MCP and manifest tools use the same declaration with runtime JSON Schema.
`Tool.Definition` is a transparent structural value with exactly one executor. Effect schemas and schemas implementing both Standard Schema V1 and Standard JSON Schema V1 are accepted. The Tool module derives model definitions and interprets invocations for the registry; callers normally rely on `Tool.make` inference rather than naming the declaration type.
Standard input schemas validate model input into the handler value. Standard output schemas validate the handler result into the model-facing value. Effect codecs retain their native decode-input and encode-output directions.
Input and output codecs are self-contained. Schema conversion cannot require services. Tool dependencies are acquired during construction and captured by `execute`.
`assistantMessageID` is the durable ID of the assistant message containing the call. The Session runner owns this association and supplies the complete context to the registry; the registry does not infer it.
Decoded tool input is passed separately to `execute`. Raw provider input and domain services do not belong in the invocation context.
Effect interruption is the cancellation mechanism. Tools may translate expected typed failures into `ToolFailure`, but must not translate interruption or defects into model-visible failures.
The record key is the authored name. Registration normalizes it before deriving the effective model-facing name. A reusable tool value has no intrinsic name.
A Location plugin receives only the narrow `Tools` registration capability, not the internal registry. Each activation acquires the Location's services, constructs its tools, and registers them in a fresh plugin-owned Scope.
Built-ins use the same tool API while capturing trusted Location services:
```ts
const filesystem = yield * FileSystem.Service
const permission = yield * PermissionV2.Service
const tools = yield * Tools.Service
yield *
tools.register({
grep: Tool.make({
description: "Search file contents",
input: Input,
output: Output,
execute: (input, context) =>
Effect.gen(function* () {
const root = yield* filesystem.resolveRoot(input)
yield* permission.assert({
sessionID: context.sessionID,
agent: context.agent,
source: {
type: "tool",
messageID: context.assistantMessageID,
callID: context.toolCallID,
},
action: "grep",
resources: [input.pattern],
save: ["*"],
metadata: { root: root.resource },
})
return yield* filesystem.grep(input, root)
}).pipe(/* translate expected typed errors to ToolFailure */),
}),
})
```
Trusted tools formulate and sequence permission requests. `PermissionV2` evaluates policy and manages approval. The registry does not inject an `assertPermission` helper.
Sharing a tool type does not imply equal authority. Built-ins and trusted Location plugins may capture services that are not available to application tools.
Invalid input never invokes the tool. Invalid output never produces a successful settlement.
`toModelOutput` is pure and total. When omitted, the encoded output remains structured output; an encoded string is also projected as text. Projection does not receive invocation identity because presentation depends only on validated input and output.
Each model request captures the effective registered `Tool` value for every advertised name. Settlement executes those captured values; later registration changes affect later requests.
Producers may cap capture or spool data before a complete tool result exists. For example, a process tool may retain output it cannot keep in memory. Producer limits must report their own loss accurately; they are separate from registry bounding and cannot claim to reconstruct bytes already discarded.
After projection, the registry bounds the channel sent to the provider. When content exists, only its textual parts are measured; structured metadata is retained unchanged without being double-counted, and native media remains unchanged under producer-owned limits. When content is empty, the structured output is measured. Oversized provider-facing text or structured output is retained in managed storage and replaced with a bounded text preview while structured metadata and media are preserved; if complete retention fails, settlement fails operationally rather than publishing lossy success. Managed paths never appear in `Tool.make`, tool output schemas, or projection callbacks solely for retention bookkeeping.
`execute.after` hooks receive the bounded settlement and its internal managed paths. Hooks may deliberately transform that settlement; the registry does not apply a second bounding pass afterward.
Leaf tools translate only errors they deliberately classify as recoverable. Broad cause-catching around an executor is invalid because it consumes interruption and defects.
## Laws
- **Single executor:** `Tool.make(config)` can invoke only `config.execute`.
