mirror of
https://github.com/windmill-labs/windmill.git
synced 2026-09-08 16:03:27 +00:00
* feat(frontend): run a linked agent's draft when testing a flow, and offer to deploy it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): settle an agent's autosave before reading it, and refresh its card on a draft save Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): deploy the agent draft that was validated, and make the draft-tools flag explicit Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): refuse a stale agent deploy, and warn when a never-deployed agent is kept as a draft Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * docs: record what inlining an agent draft puts in a preview job Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): name the draft-changes dialog after what it lists Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): refuse a draft deploy when the draft row is gone Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): apply the missing-draft refusal to raw apps too Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): stop reading a deployed resource row as a draft on deploy Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): do not mistake an outage or a vanished draft for a deploy Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): refuse an agent read whose pending draft save failed Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): space the trigger badges and right-align the agent actions Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): surface a failed agent-draft read instead of dropping it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): give the agent draft delete a baseline so a newer edit survives Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): drop the agent draft cell locally instead of deleting twice Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * test(frontend): pass the withDraft flag the guard tests were missing Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * refactor(frontend): deploy agent drafts the way Review & Deploy does Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): give the read-only flow graph its own linked-tools bucket Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): base the resource draft delete on the read that promoted it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): refresh every step linking an agent when its draft is saved Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * fix(frontend): write nothing at all when a resource draft has gone Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp * test(frontend): pin that the resource draft delete follows its baseline seed Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017B4omp8dRgmLitbpQEqFMp --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
336 lines
9.8 KiB
TypeScript
336 lines
9.8 KiB
TypeScript
import {
|
|
JobService,
|
|
type Flow,
|
|
type FlowModule,
|
|
type InputTransform,
|
|
type Job,
|
|
type RestartedFrom,
|
|
type OpenFlow,
|
|
type MemoryConfig,
|
|
type FlowValue,
|
|
type Retry
|
|
} from '$lib/gen'
|
|
import { workspaceStore } from '$lib/stores'
|
|
import { cleanExpr, emptySchema } from '$lib/utils'
|
|
import { unescapeTemplateBackticks } from '$lib/utils/templateLiteral'
|
|
import { get } from 'svelte/store'
|
|
import type { FlowModuleState } from './flowState'
|
|
import { type PickableProperties, dfs } from './previousResults'
|
|
import { forEachFlowModule } from './dfs'
|
|
import { withAgentDrafts } from './linkedAgentDrafts'
|
|
import { NEVER_TESTED_THIS_FAR } from './models'
|
|
import { sendUserToast } from '$lib/toast'
|
|
import type { ExtendedOpenFlow } from './types'
|
|
import type { GraphModuleState } from '../graph'
|
|
import type { ModulesTestStates } from '../modulesTest.svelte'
|
|
|
|
function create_context_function_template(eval_string: string, context: Record<string, any>) {
|
|
return `
|
|
return function (context) {
|
|
"use strict";
|
|
${
|
|
Object.keys(context).length > 0
|
|
? `let ${Object.keys(context).map((key) => ` ${key} = context['${key}']`)};`
|
|
: ``
|
|
}
|
|
return ${eval_string}
|
|
}`
|
|
}
|
|
|
|
function make_context_evaluator(eval_string, context): (context) => any {
|
|
let template = create_context_function_template(eval_string, context)
|
|
let functor = Function(template)
|
|
return functor()
|
|
}
|
|
|
|
export function evalValue(
|
|
k: string,
|
|
mod: FlowModule,
|
|
pickableProperties: PickableProperties | undefined,
|
|
showError: boolean
|
|
): any {
|
|
let inputTransforms = (mod.value['input_transforms'] ?? {}) as Record<string, InputTransform>
|
|
let v: any
|
|
let t = inputTransforms?.[k]
|
|
|
|
if (t?.type == 'static') {
|
|
v = t.value
|
|
} else if (t?.type == 'javascript') {
|
|
try {
|
|
let context = {
|
|
flow_input: pickableProperties?.flow_input,
|
|
results: pickableProperties?.priorIds
|
|
}
|
|
v = make_context_evaluator(t.expr, context)(context)
|
|
} catch (e) {
|
|
if (showError) {
|
|
sendUserToast(`Error evaluating ${k}: ${e.message}`, true)
|
|
}
|
|
v = undefined
|
|
}
|
|
} else {
|
|
v = undefined
|
|
}
|
|
if (v === NEVER_TESTED_THIS_FAR) {
|
|
v = undefined
|
|
}
|
|
return v
|
|
}
|
|
|
|
/** Ensure modules comes first and groups last in the value object for readable YAML export. */
|
|
function reorderFlowValue(value: ExtendedOpenFlow['value']): ExtendedOpenFlow['value'] {
|
|
const { modules, groups, ...rest } = value
|
|
return { modules, ...rest, ...(groups != null ? { groups } : {}) }
|
|
}
|
|
|
|
export function filteredContentForExport(flow: ExtendedOpenFlow) {
|
|
let o = {
|
|
summary: flow.summary,
|
|
description: flow.description,
|
|
value: reorderFlowValue(flow.value),
|
|
schema: flow.schema
|
|
}
|
|
if (flow.dedicated_worker) {
|
|
o['dedicated_worker'] = flow.dedicated_worker
|
|
}
|
|
if (flow.visible_to_runner_only) {
|
|
o['visible_to_runner_only'] = flow.visible_to_runner_only
|
|
}
|
|
if (flow.on_behalf_of_email) {
|
|
o['on_behalf_of_email'] = flow.on_behalf_of_email
|
|
}
|
|
if (flow.on_behalf_of) {
|
|
o['on_behalf_of'] = flow.on_behalf_of
|
|
}
|
|
if (flow.ws_error_handler_muted) {
|
|
o['ws_error_handler_muted'] = flow.ws_error_handler_muted
|
|
}
|
|
if (flow.tag) {
|
|
o['tag'] = flow.tag
|
|
}
|
|
return o
|
|
}
|
|
|
|
import { dfs as dfsApply } from './dfs'
|
|
import { randomUUID } from '$lib/utils/uuid'
|
|
|
|
export function cleanFlow(flow: OpenFlow | any): OpenFlow & {
|
|
tag?: string
|
|
ws_error_handler_muted?: boolean
|
|
dedicated_worker?: boolean
|
|
visible_to_runner_only?: boolean
|
|
on_behalf_of_email?: string
|
|
on_behalf_of?: string
|
|
} {
|
|
const newFlow: Flow = $state.snapshot(flow)
|
|
|
|
dfsApply(newFlow.value.modules, (mod) => {
|
|
if (mod.value.type == 'rawscript' || mod.value.type == 'script') {
|
|
Object.values(mod.value.input_transforms ?? {}).forEach((inp) => {
|
|
// for now we use the value for dynamic expression when done in the static editor so we have to resort to this
|
|
if (inp.type == 'javascript') {
|
|
//@ts-ignore
|
|
inp.value = undefined
|
|
inp.expr = cleanExpr(inp.expr)
|
|
} else {
|
|
//@ts-ignore
|
|
inp.expr = undefined
|
|
}
|
|
})
|
|
}
|
|
if (mod.value.type == 'rawscript' && mod.value.assets?.length == 0) {
|
|
mod.value.assets = undefined
|
|
}
|
|
// Generate memory_id for AI agents with auto memory if not already set
|
|
// Only if chat input is not enabled, as otherwise memory id is based on conversation id
|
|
if (!newFlow.value.chat_input_enabled && mod.value.type === 'aiagent') {
|
|
const memoryTransform = mod.value.input_transforms?.memory
|
|
if (memoryTransform?.type === 'static' && memoryTransform.value) {
|
|
const memoryValue = memoryTransform.value as MemoryConfig
|
|
if (
|
|
memoryValue.kind === 'auto' &&
|
|
memoryValue.context_length &&
|
|
memoryValue.context_length > 0 &&
|
|
!memoryValue.memory_id
|
|
) {
|
|
memoryTransform.value = {
|
|
...memoryValue,
|
|
memory_id: randomUUID()
|
|
}
|
|
}
|
|
}
|
|
}
|
|
})
|
|
if (newFlow.value.concurrency_key == '') {
|
|
newFlow.value.concurrency_key = undefined
|
|
}
|
|
|
|
return newFlow
|
|
}
|
|
|
|
export function getDefaultExpr(
|
|
key: string = 'myfield',
|
|
previousModuleId: string | undefined,
|
|
previousExpr?: string
|
|
) {
|
|
return (
|
|
previousExpr ?? (previousModuleId ? `results.${previousModuleId}.${key}` : `flow_input.${key}`)
|
|
)
|
|
}
|
|
|
|
export function jobsToResults(jobs: Job[]) {
|
|
return jobs.map((job) => {
|
|
if ('result' in job) {
|
|
return job.result
|
|
} else if (Array.isArray(job)) {
|
|
return jobsToResults(job)
|
|
}
|
|
})
|
|
}
|
|
|
|
/**
|
|
* Run the flow the editor currently holds. A step linked to a saved agent runs that agent's
|
|
* unsaved draft when there is one (`withAgentDrafts`), so testing exercises what the agent editor
|
|
* is showing rather than the deployed resource — the same rule the agent editor's own test pane
|
|
* follows. The value passed in is left alone; only what goes to the server is substituted.
|
|
*/
|
|
export async function runFlowPreview(
|
|
args: Record<string, any>,
|
|
flow: OpenFlow & { tag?: string },
|
|
path: string,
|
|
restartedFrom: RestartedFrom | undefined,
|
|
conversationId?: string | undefined,
|
|
tempScriptRefs?: Record<string, string>,
|
|
// The session's acting workspace when previewing inside an AI-session flow
|
|
// editor; falls back to the navigation workspace for full-page previews.
|
|
workspace?: string
|
|
) {
|
|
const ws = workspace ?? get(workspaceStore) ?? ''
|
|
const value = await withAgentDrafts(flow.value, ws)
|
|
return await JobService.runFlowPreview({
|
|
workspace: ws,
|
|
requestBody: {
|
|
args,
|
|
value,
|
|
path: path,
|
|
tag: flow.tag,
|
|
restarted_from: restartedFrom,
|
|
temp_script_refs: tempScriptRefs
|
|
},
|
|
memoryId: conversationId
|
|
})
|
|
}
|
|
|
|
export function codeToStaticTemplate(code?: string): string | undefined {
|
|
if (!code || typeof code != 'string') return undefined
|
|
|
|
const lines = code.split('\n')
|
|
if (lines.length == 1) {
|
|
const line = lines[0].trim()
|
|
if (line[0] == '`' && line.charAt(line.length - 1) == '`') {
|
|
return unescapeTemplateBackticks(line.slice(1, line.length - 1))
|
|
} else {
|
|
return `\$\{${line}\}`
|
|
}
|
|
}
|
|
return undefined
|
|
}
|
|
|
|
export function emptyFlowModuleState(): FlowModuleState {
|
|
return {
|
|
schema: emptySchema(),
|
|
previewResult: NEVER_TESTED_THIS_FAR
|
|
}
|
|
}
|
|
|
|
// `same_worker` hands the next step directly to the worker holding `./shared`, bypassing
|
|
// `scheduled_for`: a retry delay is silently ignored, and a sleep breaks the hand-off so the
|
|
// next step can land on another worker without `./shared`. Keep the two mutually exclusive.
|
|
export const SAME_WORKER_INCOMPATIBLE_MSG =
|
|
'Retries and sleeps are not compatible with the shared directory (`Same Worker`): retry delays would be ignored and a sleep would lose the `./shared` folder.'
|
|
|
|
// Mirrors the backend's `Retry::has_attempts`: a retry with no attempt never runs, and the
|
|
// retries tab renders it as "Disabled", so it must not block anything.
|
|
function hasRetryAttempts(retry: Retry | undefined): boolean {
|
|
return (retry?.constant?.attempts ?? 0) > 0 || (retry?.exponential?.attempts ?? 0) > 0
|
|
}
|
|
|
|
export function modulesWithRetryOrSleep(flow: FlowValue): string[] {
|
|
// The failure and preprocessor modules live outside `modules` but run as regular steps on
|
|
// the same-worker hand-off. Agent tools don't: they never go through the flow scheduler.
|
|
const roots = [flow.modules, flow.failure_module, flow.preprocessor_module]
|
|
.flat()
|
|
.filter((m) => m != undefined)
|
|
const ids: string[] = []
|
|
forEachFlowModule(
|
|
roots,
|
|
(m) => {
|
|
if (hasRetryAttempts(m.retry) || m.sleep != undefined) {
|
|
ids.push(m.id)
|
|
}
|
|
},
|
|
{ skipToolNodes: true }
|
|
)
|
|
return ids
|
|
}
|
|
|
|
export function checkIfParentLoop(
|
|
flowStore: ExtendedOpenFlow,
|
|
modId: string
|
|
): { id: string; type: 'forloopflow' | 'whileloopflow' } | undefined {
|
|
const flow: ExtendedOpenFlow = JSON.parse(JSON.stringify(flowStore))
|
|
const parents = dfs(modId, flow, true)
|
|
for (const parent of parents.slice(1)) {
|
|
if (parent.value.type === 'forloopflow' || parent.value.type === 'whileloopflow') {
|
|
return { id: parent.id, type: parent.value.type }
|
|
}
|
|
}
|
|
return undefined
|
|
}
|
|
|
|
/**
|
|
* Updates moduleStates based on test job data from modulesTestStates
|
|
* Extracts job information and converts it to GraphModuleState format
|
|
* for graph rendering
|
|
*/
|
|
export function updateDerivedModuleStatesFromTestJobs(
|
|
moduleId: string | undefined,
|
|
moduleTestStates: ModulesTestStates | undefined,
|
|
moduleStates: Record<string, GraphModuleState> | undefined
|
|
) {
|
|
if (!moduleId || !moduleTestStates || !moduleStates) {
|
|
return
|
|
}
|
|
const newStates: Record<string, GraphModuleState> = {}
|
|
|
|
const testState = moduleTestStates?.states[moduleId]
|
|
if (testState) {
|
|
if (testState.testJob) {
|
|
const job = testState.testJob
|
|
|
|
// Create GraphModuleState from job data
|
|
const moduleState: GraphModuleState = {
|
|
args: job.args,
|
|
type: job.type === 'QueuedJob' ? 'InProgress' : job['success'] ? 'Success' : 'Failure',
|
|
job_id: job.id,
|
|
tag: job.tag,
|
|
duration_ms: job['duration_ms'],
|
|
started_at: job.started_at ? new Date(job.started_at).getTime() : undefined
|
|
}
|
|
|
|
newStates[moduleId] = moduleState
|
|
} else if (testState.loading) {
|
|
// If test is loading, show as InProgress
|
|
newStates[moduleId] = {
|
|
type: 'InProgress',
|
|
args: {}
|
|
}
|
|
}
|
|
}
|
|
|
|
return {
|
|
...moduleStates,
|
|
...newStates
|
|
}
|
|
}
|