Files
claude-code/packages/builtin-tools/src/tools/ExecuteTool/ExecuteTool.ts
claude-code-best 99b9c6a400 fix: ExecuteExtraTool 加 schema 预校验防止 deferred 工具崩溃
模型通过 ExecuteExtraTool 调 CronCreate 时把字段名拼成 schedule(而非
cron),raw params 透传到 validateInput,input.cron 为 undefined,触
发 parseCronExpression 的 expr.trim() 抛 TypeError。所有参数组合同
样崩溃,与具体参数无关。

三层修复:
- ExecuteTool.call 在调 validateInput 前先跑 targetTool.inputSchema.
  safeParse(鸭子类型跳过 MCP),失败时返回 formatZodValidationError
  友好消息,成功时透传 parsed data 让 .default() 生效、strictObject
  拦截多余字段。这是架构根治,覆盖所有 deferred 工具。
- CronCreateTool.validateInput 顶部加 typeof string 守卫,给模型精确
  字段错误消息。
- parseCronExpression 顶部加 typeof string 守卫,覆盖 cronToHuman 等
  所有调用者。

新增 4 个测试覆盖:cron undefined 输入返回 null、ExecuteTool schema
验证拒绝错字段名且 validateInput 不被触达、.default() 透传、MCP-like
工具跳过 schema 校验。

Co-Authored-By: glm-5.2 <zai-org@claude-code-best.win>
2026-06-18 13:59:27 +08:00

240 lines
7.3 KiB
TypeScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import { z } from 'zod/v4'
import {
buildTool,
findToolByName,
type Tool,
type ToolDef,
type ToolUseContext,
type ToolResult,
type Tools,
} from 'src/Tool.js'
import { lazySchema } from 'src/utils/lazySchema.js'
import { createUserMessage } from 'src/utils/messages.js'
import { formatZodValidationError } from 'src/utils/toolErrors.js'
import {
extractDiscoveredToolNames,
isSearchExtraToolsEnabledOptimistic,
isSearchExtraToolsToolAvailable,
} from 'src/utils/searchExtraTools.js'
import { DESCRIPTION, getPrompt } from './prompt.js'
import { EXECUTE_TOOL_NAME } from './constants.js'
import { isDeferredTool } from '../SearchExtraToolsTool/prompt.js'
export const inputSchema = lazySchema(() =>
z.object({
tool_name: z
.string()
.describe(
'The exact name of the target tool to execute (e.g., "CronCreate", "mcp__server__action")',
),
params: z
.record(z.string(), z.unknown())
.describe('The parameters to pass to the target tool'),
}),
)
type InputSchema = ReturnType<typeof inputSchema>
export const outputSchema = lazySchema(() =>
z.object({
result: z.unknown(),
tool_name: z.string(),
}),
)
type OutputSchema = ReturnType<typeof outputSchema>
export type Output = z.infer<OutputSchema>
export const ExecuteTool = buildTool({
name: EXECUTE_TOOL_NAME,
searchHint: 'execute run invoke call a deferred tool by name with parameters',
maxResultSizeChars: 100_000,
isConcurrencySafe() {
return false
},
get inputSchema(): InputSchema {
return inputSchema()
},
get outputSchema(): OutputSchema {
return outputSchema()
},
async description() {
return DESCRIPTION
},
async prompt() {
return getPrompt()
},
async call(input, context, canUseTool, parentMessage, onProgress) {
const tools: Tools = context.options.tools ?? []
const targetTool = findToolByName(tools, input.tool_name)
if (!targetTool) {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: `Tool "${input.tool_name}" not found. Use SearchExtraTools to discover available tools.`,
}),
],
}
}
// Guard: block execution of undiscovered deferred tools.
// When tool search is active, deferred tools must be discovered via
// SearchExtraTools first so the model has seen their schemas and knows
// the correct parameters. Executing an undiscovered tool almost always
// fails with parameter validation errors.
if (
isSearchExtraToolsEnabledOptimistic() &&
isSearchExtraToolsToolAvailable(tools) &&
isDeferredTool(targetTool)
) {
const discovered = extractDiscoveredToolNames(context.messages)
if (!discovered.has(input.tool_name)) {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: `Tool "${input.tool_name}" has not been discovered yet. You must first use SearchExtraTools to discover this tool before executing it.\n\nUsage: SearchExtraTools("select:${input.tool_name}")`,
}),
],
}
}
}
// Check if the target tool is currently enabled
if (!targetTool.isEnabled()) {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: `工具 "${input.tool_name}" 当前不可用Remote Control 未连接。`,
}),
],
}
}
// Schema-validate params against the target tool BEFORE delegating.
// ExecuteExtraTool passes raw params straight from the model to
// validateInput/call without re-running the target's zod schema, so a
// wrong field name (e.g. 'schedule' instead of 'cron') or a missing
// required field reaches the tool as undefined and the first
// .trim()/.length/.split() crashes with "undefined is not an object".
// CronCreateTool's .trim() crash was the reported symptom; centralizing
// the check here covers every deferred tool without relying on each one
// to defensively guard its own validateInput. Duck-typed so MCP tools
// (whose schema is inputJSONSchema, not zod) skip this branch.
const targetSchema = targetTool.inputSchema as
| { safeParse?: (data: unknown) => unknown }
| undefined
if (targetSchema?.safeParse) {
const parsed = targetSchema.safeParse(input.params) as
| { success: true; data: Record<string, unknown> }
| { success: false; error: z.ZodError }
if (!parsed.success) {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: formatZodValidationError(input.tool_name, parsed.error),
}),
],
}
}
// Use parsed params going forward — picks up .default() values and
// strips unknown keys for strictObject schemas so validateInput/call
// never see fields they don't expect.
input.params = parsed.data
}
// Validate input before delegating — prevents crashes when the model
// omits required params (e.g. TeamCreate without team_name →
// sanitizeName(undefined).replace() TypeError).
if (targetTool.validateInput) {
const validation = await targetTool.validateInput(
input.params as Record<string, unknown>,
context,
)
if (!validation.result) {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: `Invalid parameters for tool "${input.tool_name}": ${validation.message}`,
}),
],
}
}
}
// Check permissions on the target tool
const permResult = await targetTool.checkPermissions?.(
input.params as Record<string, unknown>,
context,
)
if (permResult && permResult.behavior === 'deny') {
return {
data: {
result: null,
tool_name: input.tool_name,
},
newMessages: [
createUserMessage({
content: `Permission denied for tool "${input.tool_name}": ${permResult.message ?? 'Permission denied'}`,
}),
],
}
}
// Delegate execution to the target tool
const targetResult: ToolResult<unknown> = await targetTool.call(
input.params as Record<string, unknown>,
context,
canUseTool,
parentMessage,
onProgress,
)
return {
...targetResult,
data: {
result: targetResult.data,
tool_name: input.tool_name,
},
}
},
async checkPermissions() {
return {
behavior: 'passthrough',
message: 'ExecuteExtraTool delegates permission to the target tool.',
}
},
renderToolUseMessage(input) {
return `${input.tool_name}`
},
userFacingName() {
return 'ExecuteExtraTool'
},
mapToolResultToToolResultBlockParam(content, toolUseID) {
return {
tool_use_id: toolUseID,
type: 'tool_result',
content: JSON.stringify(content),
}
},
} satisfies ToolDef<InputSchema, Output>)