opensource/sim
3
0
Fork 0
mirror of https://github.com/simstudioai/sim.git synced 2026-09-24 15:45:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

improvement(provenance): name every guard that can latch a registry (#6513)

Browse Source 
* fix(provenance): name every guard that can latch a registry

A production latch reported `reason: "unspecified"` because 44 call sites took
the default. The reason is the only thing that names which guard tripped, and a
refusal surfaces many frames later as one fixed sentence, so an unnamed latch is
undiagnosable — that is what left an incident's origin unidentified for a day.

Give each call site a literal that names its guard, add the 20 new literals to
the reason union, and sort them into the existing error/warn split: a guard that
should not trip on a healthy run reports at error, everything else stays at warn.
`log-creation-skipped` joins the by-design set since it fires on every run that
does not persist a log.

Make `reason` required on both `markIncomplete` and `markInputPathIncomplete`, so
omission is a compile error rather than a silent `unspecified`. A caller with
genuinely nothing to say now passes `'unspecified'` where a reviewer can see it.

The three remaining bare calls are on ResolvedSecretTraceProvenanceAccumulator,
a different class with no reason concept.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

* test(provenance): pin the new reasons and the guard that latched in production

Cover what the reason set is for rather than only that it compiles: the
non-enumerable tool-params branch now asserts it names `tool-input-not-enumerable`,
which is the guard the production logs showed reporting `unspecified`, and every
new literal asserts which stream it reports on — error for a guard that cannot
trip on a healthy run, warn for one reachable without a fault, silent for the
by-design log-less session.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

* refactor(provenance): make reason the only locator, not one of two

Two fields had grown into competing answers to the same question. `origin` is a
free-form label for which importer accepted an already-incomplete bundle; four
latches had started passing `markIncomplete('unspecified', { origin })`, using it
to stand in for a reason that did not exist yet. That splits one fact across a
closed enum and an open string, leaving neither worth alerting on.

Give those four the literal they were reaching for — none needed a new one — and
split the five reasons that covered genuinely different guards, so the reason
alone locates the site rather than needing an origin beside it. `origin` keeps its
narrow job, now documented: it disambiguates importers that share one guard, and a
latch that wants an origin because no reason fits should add a reason instead.

No production call site passes 'unspecified' any more.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

* fix(provenance): stop two expected states from reporting as faults

Cursor Bugbot caught `backfill-checkpoint-*`: the guard covered four conditions
under one reason classified as an originating fault, and one of them — a state
persisted before the checkpoint contract existed — is what essentially every
legacy row looks like. A backfill over historical rows would have put one error
line per row into the stream the error/warn split exists to protect.

Auditing the rest of the error-level reasons for the same shape found a second:
a client tool invoked without a run id has no binding to unseal against, so it
took the `[null, null]` path and reported `client-tool-seal-failed` at error on an
ordinary configuration.

Split both along the line that matters — absent versus unusable, not attempted
versus failed — and classify each half: expected states warn, genuine faults keep
error. `backfill-scope-mismatch` is retired; it named one of its four conditions.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Vikhyath Mondreti
2026-08-10 18:09:39 -07:00
committed by GitHub
co-authored by Claude Opus 5
parent 5f4dc195fb
commit 4bafd14110
33 changed files with 253 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files
apps/sim/app/api/knowledge/search/route.ts
+2 -2
View File
@@ -381,7 +381,7 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
results,
})
if (!resultProvenanceSnapshot.imported) {
resultSecretRegistry.markIncomplete()
resultSecretRegistry.markIncomplete('knowledge-result-provenance-unavailable')
if (useReranker) {
return NextResponse.json(
{ error: 'Knowledge result secret provenance is unavailable' },
@@ -608,7 +608,7 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
'knowledge'
))
) {
resultSecretRegistry.markIncomplete()
resultSecretRegistry.markIncomplete('knowledge-result-provenance-unavailable')
}
}
apps/sim/app/api/knowledge/secret-provenance.ts
+1 -1
View File
@@ -154,7 +154,7 @@ export async function createKnowledgeProvenanceResponse(options: {
})
for (const provenance of options.provenances) {
if (provenance.status === 'unknown') {
registry.markIncomplete()
registry.markIncomplete('durable-provenance-unknown')
break
}
const sourceRegistry = await createDurableSecretProvenanceRegistry(provenance, {
apps/sim/app/api/memory/secret-provenance.ts
+2 -2
View File
@@ -98,7 +98,7 @@ export async function createMemoryResponse(options: {
workspaceId: options.workspaceId,
})
if (options.memories.length > MAX_PRIVATE_MEMORY_CROSSINGS) {
registry.markIncomplete()
registry.markIncomplete('memory-crossing-capacity-exceeded')
} else {
const ids = [...new Set(options.memories.map((record) => record.id))]
const memoriesById = new Map<string, MemoryCrossing[]>()
@@ -123,7 +123,7 @@ export async function createMemoryResponse(options: {
provenanceEntryCount > MAX_PRIVATE_MEMORY_PROVENANCE_ENTRIES ||
provenanceBytes > MAX_PRIVATE_MEMORY_PROVENANCE_BYTES
) {
registry.markIncomplete()
registry.markIncomplete('memory-crossing-capacity-exceeded')
break
}
const sidecarById = new Map(sidecars.map((sidecar) => [sidecar.memoryId, sidecar]))
apps/sim/app/api/workflows/[id]/log/route.ts
+1 -1
View File
@@ -142,7 +142,7 @@ export const POST = withRouteHandler(
: undefined
const trustedProvenance = trustedExecutionState?.resolvedSecretTraceProvenance
if (trustedProvenance === undefined) {
resolvedSecretTraceRegistry.markIncomplete()
resolvedSecretTraceRegistry.markIncomplete('restored-provenance-untrusted')
} else {
await resolvedSecretTraceRegistry.importProvenance(trustedProvenance, {
trusted: true,
apps/sim/executor/handlers/agent/memory.test.ts
+2 -2
View File
@@ -278,7 +278,7 @@ describe('Memory', () => {
it('persists raw memory with unknown lineage when provenance is unavailable', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const appendMessage = vi
.spyOn(memoryService as any, 'appendMessage')
.mockResolvedValue(undefined)
@@ -293,7 +293,7 @@ describe('Memory', () => {
it('seeds raw memory with unknown lineage when provenance is unavailable', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const seedMemoryRecord = vi
.spyOn(memoryService as any, 'seedMemoryRecord')
.mockResolvedValue(undefined)
apps/sim/executor/handlers/generic/generic-handler.ts
+3 -2
View File
@@ -199,7 +199,8 @@ export class GenericBlockHandler implements BlockHandler {
boundary && boundary.paths.length > 0 && registry?.hasResolvedInputProjections()
? registry.projectResolvedInputSelections(inputs)
: undefined
if (projectedInputs?.complete === false) registry?.markIncomplete()
if (projectedInputs?.complete === false)
registry?.markIncomplete('structural-input-projection-incomplete')
if (projectedInputs?.complete && boundary && tool && registry) {
for (const projection of projectedInputs.values) {
@@ -233,7 +234,7 @@ export class GenericBlockHandler implements BlockHandler {
continue
}
if (boundary.requiredProjectionRoots.has(projection.path[0])) {
registry.markIncomplete()
registry.markIncomplete('structural-input-root-unprojected')
}
continue
}
apps/sim/executor/handlers/mothership/mothership-handler.ts
+6 -6
View File
@@ -372,7 +372,7 @@ async function consumeMothershipProvenance(
return false
}
if (inspection.status === 'invalid') {
registry?.markIncomplete()
registry?.markIncomplete('mothership-provenance-invalid')
throw new Error('Mothership response provenance metadata is invalid')
}
@@ -399,7 +399,7 @@ function inspectMothershipResponseCapability(
return false
}
registry?.markIncomplete()
registry?.markIncomplete('mothership-provenance-invalid')
throw new Error('Mothership response provenance metadata is invalid')
}
@@ -464,7 +464,7 @@ async function readMothershipExecuteResponse(
result = (await response.json()) as MothershipExecuteResult
} catch (error) {
if (expectsProvenance) {
registry?.markIncomplete()
registry?.markIncomplete('mothership-response-unreadable')
throw new Error('Mothership response provenance metadata is invalid')
}
throw error
@@ -528,7 +528,7 @@ async function readMothershipExecuteResponse(
return finalResult
} finally {
if (expectsProvenance && !finalResult && !receivedTerminalProvenance) {
registry?.markIncomplete()
registry?.markIncomplete('mothership-provenance-missing')
}
reader.releaseLock()
}
@@ -630,7 +630,7 @@ function createMothershipStreamingExecution(
}
} finally {
if (expectsProvenance && !sawFinal && !receivedTerminalProvenance) {
options.registry?.markIncomplete()
options.registry?.markIncomplete('mothership-provenance-missing')
}
cleanup()
reader?.releaseLock()
@@ -948,7 +948,7 @@ export class MothershipBlockHandler implements BlockHandler {
try {
payload = (await response.clone().json()) as MothershipExecuteResult
} catch {
resultRegistry?.markIncomplete()
resultRegistry?.markIncomplete('mothership-response-unreadable')
throw new Error('Mothership response provenance metadata is invalid')
}
await consumeMothershipProvenance(payload, response, resultRegistry)
apps/sim/executor/handlers/pi/local/sim-tools.test.ts
+1 -1
View File
@@ -320,7 +320,7 @@ describe('buildSimToolSpecs', () => {
output: { result: 'untrusted output' },
})
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const [spec] = await buildSimToolSpecs(executionContext(registry), toolInput)
const result = await spec.execute({})
apps/sim/executor/handlers/pi/pi-handler.test.ts
+1 -1
View File
@@ -227,7 +227,7 @@ describe('PiBlockHandler', () => {
it('fails closed when task provenance is incomplete', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
await expect(
handler.execute(
apps/sim/executor/handlers/pi/search/tool.test.ts
+2 -2
View File
@@ -288,7 +288,7 @@ describe('buildPiSearchToolSpec', () => {
it('fails closed before search when provenance is incomplete', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const result = await buildTool('exa', executionContext(registry)).execute({ query: 'pi' })
@@ -303,7 +303,7 @@ describe('buildPiSearchToolSpec', () => {
const registry = new ResolvedSecretTraceRegistry()
const mergeSpy = vi.spyOn(registry, 'mergeToolCallRegistry')
mockExecuteTool.mockImplementation(async (_toolId, _params, options) => {
options.resolvedSecretTraceRegistry.markIncomplete()
options.resolvedSecretTraceRegistry.markIncomplete('unspecified')
return {
success: true,
output: {
apps/sim/executor/utils/resolved-secret-content-projection.test.ts
+3 -3
View File
@@ -70,7 +70,7 @@ describe('projectResolvedSecretModelContent', () => {
value: '{{TOKEN}}',
})
registry.markIncomplete()
registry.markIncomplete('unspecified')
expect(projectResolvedSecretModelContent('secret-value', registry)).toEqual({ safe: false })
expect(projectResolvedSecretModelContent('secret-value', undefined)).toEqual({ safe: false })
})
@@ -368,7 +368,7 @@ describe('projectResolvedSecretModelJsonContent', () => {
it('does not invoke JSON serialization when provenance is incomplete', () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const toJSON = vi.fn(() => ({ value: 'untrusted' }))
expect(projectResolvedSecretModelJsonContent({ toJSON }, registry)).toEqual({ safe: false })
@@ -474,7 +474,7 @@ describe('projectResolvedSecretDiagnosticError', () => {
it('falls back to text-free structure when provenance is missing or incomplete', () => {
const error = new Error('secret __var_API_KEY')
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
expect(projectResolvedSecretDiagnosticError(error, undefined)).toEqual({
errorType: 'error',
apps/sim/executor/utils/resolved-secret-trace-registry.test.ts
+77 -2
View File
@@ -141,7 +141,7 @@ describe('ResolvedSecretTraceProvenanceAccumulator', () => {
entries: [{ name: 'TOKEN', encryptedValue: 'encrypted-value' }],
scope,
})
accumulator.markIncomplete()
accumulator.markIncomplete('unspecified')
expect(accumulator.exportProvenance().entries).toEqual([])
})
})
@@ -1443,6 +1443,81 @@ describe('incompleteness diagnostics', () => {
expect(reasons).not.toContain('value-provenance-filter-incomplete')
})
it.each([
'tool-input-not-enumerable',
'tool-params-transform-failed',
'structural-input-projection-incomplete',
'mothership-provenance-invalid',
'client-tool-seal-failed',
'knowledge-row-missing',
'knowledge-row-content-mismatch',
'mothership-response-unreadable',
'structural-input-root-unprojected',
'backfill-checkpoint-unusable',
] as const)('reports %s at error, since it cannot trip on a healthy run', (reason) => {
new ResolvedSecretTraceRegistry([], scope).markIncomplete(reason)
expect(mockLogger.error).toHaveBeenCalledWith(
'Resolved secret registry marked incomplete',
expect.objectContaining({ reason })
)
expect(mockLogger.warn).not.toHaveBeenCalled()
})
it.each([
'mothership-provenance-missing',
'client-tool-completion-missing',
'client-tool-completion-deferred',
'client-tool-completion-unidentified',
'client-tool-execution-untrusted',
'client-tool-content-unavailable',
'knowledge-result-provenance-unavailable',
'knowledge-response-capacity-exceeded',
'memory-crossing-capacity-exceeded',
'workspace-scope-missing',
'mounted-file-provenance-unavailable',
'table-snapshot-unsafe-for-mount',
'restored-provenance-untrusted',
'backfill-checkpoint-absent',
'client-tool-seal-absent',
] as const)('reports %s at warn, since it is reachable without a fault', (reason) => {
new ResolvedSecretTraceRegistry([], scope).markIncomplete(reason)
expect(mockLogger.warn).toHaveBeenCalledWith(
'Resolved secret registry marked incomplete',
expect.objectContaining({ reason })
)
expect(mockLogger.error).not.toHaveBeenCalled()
})
/**
* The taxonomy's rule is that an error reason cannot trip on a healthy run. A backfill walking
* historical rows hits the no-checkpoint case on essentially every legacy row, so classifying it
* as a fault would put one error line per row into the stream this split exists to protect.
*/
it('separates an absent checkpoint from an unusable one, so a backfill cannot flood errors', () => {
new ResolvedSecretTraceRegistry([], scope).markIncomplete('backfill-checkpoint-absent')
expect(mockLogger.error).not.toHaveBeenCalled()
expect(mockLogger.warn).toHaveBeenCalledWith(
'Resolved secret registry marked incomplete',
expect.objectContaining({ reason: 'backfill-checkpoint-absent' })
)
vi.clearAllMocks()
new ResolvedSecretTraceRegistry([], scope).markIncomplete('backfill-checkpoint-unusable')
expect(mockLogger.error).toHaveBeenCalledWith(
'Resolved secret registry marked incomplete',
expect.objectContaining({ reason: 'backfill-checkpoint-unusable' })
)
})
it('does not report a log-less session at all, since it fires on every such run', () => {
new ResolvedSecretTraceRegistry([], scope).markIncomplete('log-creation-skipped')
expect(mockLogger.error).not.toHaveBeenCalled()
expect(mockLogger.warn).not.toHaveBeenCalled()
})
it('reports an incoming incomplete bundle at warn, since no catalog was ever on offer', () => {
const registry = new ResolvedSecretTraceRegistry([], scope)
@@ -1465,7 +1540,7 @@ describe('incompleteness diagnostics', () => {
it('keeps an unaudited caller taking the default reason out of the error stream', () => {
const registry = new ResolvedSecretTraceRegistry([], scope)
registry.markIncomplete()
registry.markIncomplete('unspecified')
expect(mockLogger.error).not.toHaveBeenCalled()
expect(mockLogger.warn).toHaveBeenCalledWith(
apps/sim/executor/utils/resolved-secret-trace-registry.ts
+61 -2
View File
@@ -41,6 +41,36 @@ export type ResolvedSecretIncompletenessReason =
| 'value-provenance-filter-incomplete'
| 'durable-provenance-unknown'
| 'durable-provenance-malformed'
| 'tool-input-not-enumerable'
| 'tool-params-transform-failed'
| 'structural-input-projection-incomplete'
| 'structural-input-root-unprojected'
| 'mothership-provenance-invalid'
| 'mothership-response-unreadable'
| 'mothership-provenance-missing'
| 'client-tool-seal-absent'
| 'client-tool-seal-failed'
| 'client-tool-completion-missing'
| 'client-tool-completion-deferred'
| 'client-tool-completion-unidentified'
| 'client-tool-execution-untrusted'
| 'client-tool-content-unavailable'
| 'knowledge-result-provenance-unavailable'
| 'knowledge-response-capacity-exceeded'
| 'knowledge-row-missing'
| 'knowledge-row-content-mismatch'
| 'memory-crossing-capacity-exceeded'
| 'workspace-scope-missing'
| 'mounted-file-provenance-unavailable'
| 'table-snapshot-unsafe-for-mount'
| 'restored-provenance-untrusted'
| 'backfill-checkpoint-absent'
| 'backfill-checkpoint-unusable'
| 'log-creation-skipped'
/**
* Only for a caller that has not been given a reason yet. A refusal reporting this names no
* guard, which is the state that made a production latch untraceable — prefer adding a literal.
*/
| 'unspecified'
/**
@@ -68,6 +98,16 @@ const ORIGINATING_FAULT_REASONS = new Set<ResolvedSecretIncompletenessReason>([
'value-provenance-untrusted',
'value-provenance-import-failed',
'durable-provenance-malformed',
'tool-input-not-enumerable',
'tool-params-transform-failed',
'structural-input-projection-incomplete',
'mothership-provenance-invalid',
'client-tool-seal-failed',
'knowledge-row-missing',
'knowledge-row-content-mismatch',
'mothership-response-unreadable',
'structural-input-root-unprojected',
'backfill-checkpoint-unusable',
])
/**
@@ -77,6 +117,8 @@ const ORIGINATING_FAULT_REASONS = new Set<ResolvedSecretIncompletenessReason>([
*/
const BY_DESIGN_INCOMPLETENESS_REASONS = new Set<ResolvedSecretIncompletenessReason>([
'constructed-incomplete',
/** A session that will not persist a log has nothing to vouch for; it fires on every such run. */
'log-creation-skipped',
])
/**
@@ -188,6 +230,16 @@ type PreparedProvenanceFilterResult =
/** Extra attribution for a latch: which registry it propagated from, and which importer caused it. */
interface MarkIncompleteContext {
source?: ResolvedSecretTraceRegistry
/**
* Which importer accepted an already-incomplete bundle — only meaningful where several callers
* share one guard, as {@link ImportResolvedSecretTraceProvenanceOptions.origin} describes.
*
* It is not a second way to say what `reason` says. A latch that reaches for an origin because no
* reason fits is the signal to add a reason literal instead: `reason` is a closed set that can be
* alerted on and aggregated, and splitting the same fact across two fields leaves neither
* trustworthy. Passing `'unspecified'` alongside an origin is the shape that produced a
* production latch naming no guard at all.
*/
origin?: string
}
@@ -1418,8 +1470,15 @@ export class ResolvedSecretTraceRegistry {
return !this.complete || this.incompleteInputPaths.size > 0
}
/**
* `reason` is required. It defaulted to `'unspecified'`, and every caller that took the default
* produced a latch naming no guard — which is exactly the state that made a production incident
* untraceable for a day. Making omission a compile error is what keeps the reason set honest as
* new guards are added; a caller that genuinely has nothing to say passes `'unspecified'` on
* purpose, where a reviewer can see it.
*/
markIncomplete(
reason: ResolvedSecretIncompletenessReason = 'unspecified',
reason: ResolvedSecretIncompletenessReason,
context: MarkIncompleteContext = {}
): void {
if (context.source) this.inheritIncompletenessReasonsFrom(context.source)
@@ -1870,7 +1929,7 @@ export class ResolvedSecretTraceRegistry {
private markInputPathIncomplete(
path: ResolvedSecretInputPath | undefined,
reason: ResolvedSecretIncompletenessReason = 'unspecified',
reason: ResolvedSecretIncompletenessReason,
origin?: string
): void {
if (!path || path.length === 0) {
apps/sim/lib/copilot/request/handlers/tool.ts
+1 -1
View File
@@ -235,7 +235,7 @@ export async function prePersistClientExecutableToolCall(
toolInput: data.arguments,
})
} catch (error) {
execContext.resolvedSecretTraceRegistry.markIncomplete()
execContext.resolvedSecretTraceRegistry.markIncomplete('client-tool-seal-failed')
logger.warn('Failed to seal client tool provenance', {
toolCallId: data.toolCallId,
error: getErrorMessage(error),
apps/sim/lib/copilot/request/lifecycle/run.test.ts
+1 -1
View File
@@ -950,7 +950,7 @@ describe('runCopilotLifecycle', () => {
it('does not block ordinary initial Go payloads on unrelated incomplete provenance', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const result = await runCopilotLifecycle(
{ message: 'possibly secret', messageId: 'stream-incomplete-projection' },
apps/sim/lib/copilot/request/tools/client.ts
+22 -11
View File
@@ -91,8 +91,14 @@ export async function waitForClientToolCompletion({
const finishPendingActivation = toolRegistry?.beginPendingActivation()
let content: Awaited<ReturnType<typeof unsealClientToolCompletion>> = null
try {
/**
* A tool invoked without a run id has no binding to unseal against, which is a configuration
* rather than a fault. Tracking whether unsealing was even attempted keeps that ordinary case
* out of the error stream while a genuine unseal failure stays in it.
*/
const sealingAttempted = Boolean(binding && registry && toolRegistry && registryCanImport)
const [sealedContent, sealedContext] =
binding && registry && toolRegistry && registryCanImport
sealingAttempted && binding && registry
? await Promise.all([
unsealClientToolCompletion(completion.data, binding),
unsealClientToolContext(completion.data, binding, registry),
@@ -100,7 +106,9 @@ export async function waitForClientToolCompletion({
: [null, null]
if (toolRegistry && registryCanImport) {
if (!sealedContent || !sealedContext) {
toolRegistry.markIncomplete()
toolRegistry.markIncomplete(
sealingAttempted ? 'client-tool-seal-failed' : 'client-tool-seal-absent'
)
} else {
const imported = await toolRegistry.importProvenance(sealedContext.provenance, {
origin: 'copilotToolClient.sealedContext',
@@ -116,7 +124,9 @@ export async function waitForClientToolCompletion({
}
}
} catch {
toolRegistry?.markIncomplete('unspecified', { origin: 'copilotToolClient.sealedContext' })
toolRegistry?.markIncomplete('client-tool-seal-failed', {
origin: 'copilotToolClient.sealedContext',
})
} finally {
finishPendingActivation?.()
}
@@ -243,18 +253,18 @@ export async function waitForWorkflowToolCompletion({
try {
completion = await waitForToolCompletion(toolCallId, timeoutMs, abortSignal)
if (!completion) {
toolRegistry?.markIncomplete()
toolRegistry?.markIncomplete('client-tool-completion-missing')
return null
}
const executionId = getWorkflowToolCompletionExecutionId(completion.data)
const deploymentError = getAsyncWorkflowDeploymentError(completion.data)
if (completion.status === ASYNC_TOOL_CONFIRMATION_STATUS.background) {
toolRegistry?.markIncomplete()
toolRegistry?.markIncomplete('client-tool-completion-deferred')
return structuralWorkflowCompletion(completion.status, workflowId, executionId)
}
if (!workflowId || !executionId) {
toolRegistry?.markIncomplete()
toolRegistry?.markIncomplete('client-tool-completion-unidentified')
const structuralStatus =
completion.status === MothershipStreamV1ToolOutcome.success
? MothershipStreamV1ToolOutcome.error
@@ -279,12 +289,12 @@ export async function waitForWorkflowToolCompletion({
}
if (!trustedExecution) {
toolRegistry?.markIncomplete()
toolRegistry?.markIncomplete('client-tool-execution-untrusted')
return structuralWorkflowCompletion(completion.status, workflowId, executionId)
}
if (!trustedExecution.contentAvailable) {
toolRegistry?.markIncomplete()
toolRegistry?.markIncomplete('client-tool-content-unavailable')
return structuralWorkflowCompletion(
getWorkflowToolConfirmationStatus(trustedExecution.status),
workflowId,
@@ -297,7 +307,8 @@ export async function waitForWorkflowToolCompletion({
toolRegistry.isPermanentlyIncomplete() ||
!trustedExecution.provenance.complete
) {
if (!trustedExecution.provenance.complete) toolRegistry?.markIncomplete()
if (!trustedExecution.provenance.complete)
toolRegistry?.markIncomplete('source-provenance-incomplete')
return structuralWorkflowCompletion(
getWorkflowToolConfirmationStatus(trustedExecution.status),
workflowId,
@@ -317,9 +328,9 @@ export async function waitForWorkflowToolCompletion({
},
{ trusted: true }
)
if (!imported) toolRegistry.markIncomplete()
if (!imported) toolRegistry.markIncomplete('value-provenance-import-failed')
} catch (error) {
toolRegistry.markIncomplete()
toolRegistry.markIncomplete('value-provenance-import-failed')
logger.warn('Failed to import bound workflow provenance', {
toolCallId,
workflowId,
apps/sim/lib/copilot/request/tools/executor.test.ts
+3 -3
View File
@@ -237,7 +237,7 @@ describe('executeToolAndReport provenance isolation', () => {
_params: Record<string, unknown>,
toolContext: ExecutionContext
) => {
toolContext.resolvedSecretTraceRegistry?.markIncomplete()
toolContext.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
return { success: true, output: { value: 'secret-value' } }
}
)
@@ -270,7 +270,7 @@ describe('executeToolAndReport provenance isolation', () => {
_params: Record<string, unknown>,
toolContext: ExecutionContext
) => {
toolContext.resolvedSecretTraceRegistry?.markIncomplete()
toolContext.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
throw new Error('secret-value')
}
)
@@ -300,7 +300,7 @@ describe('executeToolAndReport provenance isolation', () => {
_params: Record<string, unknown>,
toolContext: ExecutionContext
) => {
toolContext.resolvedSecretTraceRegistry?.markIncomplete()
toolContext.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
abortController.abort()
return { success: true, output: { value: 'secret-value' } }
}
apps/sim/lib/copilot/request/tools/resolved-secret-result.test.ts
+1 -1
View File
@@ -335,7 +335,7 @@ describe('projectToolResultForCopilot', () => {
'incomplete',
(() => {
const registry = createRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
return registry
})(),
],
apps/sim/lib/copilot/request/tools/tables.test.ts
+2 -2
View File
@@ -379,7 +379,7 @@ describe('maybeWriteOutputToTable', () => {
it('persists raw rows with unknown provenance when lineage is incomplete', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const result = await maybeWriteOutputToTable(
FunctionExecute.id,
@@ -678,7 +678,7 @@ describe('maybeWriteReadCsvToTable', () => {
it('imports raw CSV rows with unknown provenance when lineage is incomplete', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const result = await maybeWriteReadCsvToTable(
ReadTool.id,
apps/sim/lib/copilot/tools/handlers/function-execute.ts
+12 -7
View File
@@ -118,9 +118,9 @@ async function importMountedWorkspaceFileProvenance(args: {
},
registry: args.registry,
})
if (!imported) args.registry.markIncomplete()
if (!imported) args.registry.markIncomplete('mounted-file-provenance-unavailable')
} catch {
args.registry.markIncomplete()
args.registry.markIncomplete('mounted-file-provenance-unavailable')
}
}
@@ -464,9 +464,10 @@ export async function resolveInputFiles(
workspaceId,
rowsVersion: snapshot.version,
})
if (!safeForModelMount) resolvedSecretTraceRegistry.markIncomplete()
if (!safeForModelMount)
resolvedSecretTraceRegistry.markIncomplete('table-snapshot-unsafe-for-mount')
} catch {
resolvedSecretTraceRegistry.markIncomplete()
resolvedSecretTraceRegistry.markIncomplete('table-snapshot-unsafe-for-mount')
}
if (hasCloudStorage()) {
@@ -538,7 +539,7 @@ export async function resolveInputFiles(
})
}
} catch {
resolvedSecretTraceRegistry.markIncomplete('unspecified', {
resolvedSecretTraceRegistry.markIncomplete('source-provenance-incomplete', {
origin: 'copilotFunctionExecute.result',
})
}
@@ -572,9 +573,13 @@ async function importMountedProvenance(
trusted: true,
})
if (!imported)
target.markIncomplete('unspecified', { origin: 'copilotFunctionExecute.crossing' })
target.markIncomplete('value-provenance-import-failed', {
origin: 'copilotFunctionExecute.crossing',
})
} catch {
target.markIncomplete('unspecified', { origin: 'copilotFunctionExecute.crossing' })
target.markIncomplete('value-provenance-import-failed', {
origin: 'copilotFunctionExecute.crossing',
})
}
}
apps/sim/lib/copilot/tools/server/knowledge/knowledge-base.ts
+1 -1
View File
@@ -296,7 +296,7 @@ export const knowledgeBaseServerTool: BaseServerTool<KnowledgeBaseArgs, Knowledg
results,
})
if (!resultProvenance.imported) {
resultRegistry.markIncomplete()
resultRegistry.markIncomplete('knowledge-result-provenance-unavailable')
throw new Error('Knowledge result secret provenance is unavailable')
}
apps/sim/lib/copilot/tools/server/table/user-table.ts
+1 -1
View File
@@ -427,7 +427,7 @@ async function importRowsForModel(
const registry = context.resolvedSecretTraceRegistry
if (!registry) return
if (!context.workspaceId) {
registry.markIncomplete()
registry.markIncomplete('workspace-scope-missing')
return
}
apps/sim/lib/knowledge/model-input-provenance.test.ts
+1 -1
View File
@@ -216,7 +216,7 @@ describe('Knowledge model input provenance', () => {
it('fails before model egress when an active request registry is incomplete', () => {
const registry = new ResolvedSecretTraceRegistry([])
registry.markIncomplete()
registry.markIncomplete('unspecified')
expect(() =>
runWithKnowledgeModelInputProvenance(registry, () =>
apps/sim/lib/knowledge/secret-provenance.ts
+5 -5
View File
@@ -458,7 +458,7 @@ export async function importKnowledgePersistedResponseSecretProvenance(options:
documents.length > MAX_KNOWLEDGE_RESPONSE_PROVENANCE_ROWS ||
chunks.length > MAX_KNOWLEDGE_RESPONSE_PROVENANCE_ROWS
) {
options.registry.markIncomplete()
options.registry.markIncomplete('knowledge-response-capacity-exceeded')
return false
}
@@ -480,14 +480,14 @@ export async function importKnowledgePersistedResponseSecretProvenance(options:
const documentById = new Map(documentRows.map((row) => [row.id, row]))
const chunkById = new Map(chunkRows.map((row) => [row.id, row]))
if (documentById.size !== documentIds.length || chunkById.size !== chunkIds.length) {
options.registry.markIncomplete()
options.registry.markIncomplete('knowledge-row-missing')
return false
}
for (const item of documents) {
const row = documentById.get(item.id)
if (!row) {
options.registry.markIncomplete()
options.registry.markIncomplete('knowledge-row-missing')
return false
}
const source = createKnowledgeDocumentSourceValue(row)
@@ -496,7 +496,7 @@ export async function importKnowledgePersistedResponseSecretProvenance(options:
createKnowledgeDocumentSourceValue(item.source)
)
if (!actualSourceHash || !expectedSourceHash || actualSourceHash !== expectedSourceHash) {
options.registry.markIncomplete()
options.registry.markIncomplete('knowledge-row-content-mismatch')
return false
}
const provenance = filterKnowledgeDocumentMetadataSecretProvenance(
@@ -513,7 +513,7 @@ export async function importKnowledgePersistedResponseSecretProvenance(options:
for (const item of chunks) {
const row = chunkById.get(item.id)
if (!row || row.documentId !== item.documentId || row.content !== item.content) {
options.registry.markIncomplete()
options.registry.markIncomplete('knowledge-row-content-mismatch')
return false
}
const provenance = readBoundKnowledgeEmbeddingSecretProvenance(row)
apps/sim/lib/logs/execution/logging-session.ts
+2 -2
View File
@@ -311,7 +311,7 @@ export class LoggingSession {
if (!isResolvedSecretTraceProvenanceV1(provenance)) {
const incomplete = new ResolvedSecretTraceRegistry()
incomplete.markIncomplete()
incomplete.markIncomplete('restored-provenance-untrusted')
return incomplete
}
@@ -781,7 +781,7 @@ export class LoggingSession {
[],
scopeUserId ? { userId: scopeUserId, workspaceId } : undefined
)
if (skipLogCreation) this.resolvedSecretTraceRegistry.markIncomplete()
if (skipLogCreation) this.resolvedSecretTraceRegistry.markIncomplete('log-creation-skipped')
}
try {
apps/sim/lib/model-router/resolve.test.ts
+1 -1
View File
@@ -225,7 +225,7 @@ describe('resolveAutoModel', () => {
it('does not gate caller-projected signals on ambient registry completeness', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
mockFetchGo.mockResolvedValue(routerResponse({ choice: '1' }))
const result = await resolveAutoModel({
apps/sim/lib/table/backfill-runner.ts
+12 -2
View File
@@ -72,14 +72,24 @@ export async function createBackfillExecutionSecretRegistry(options: {
? options.executionData.executionState
: undefined
const provenance = state?.resolvedSecretTraceProvenance
/**
* A state persisted before the checkpoint contract carries no version at all. That is the bulk of
* any backfill over historical rows, so it is separated from a checkpoint that exists but cannot
* be used — only the latter is worth an error, and conflating them would put one line per legacy
* row into the error stream.
*/
const checkpointPresent =
state?.resolvedSecretTraceCheckpointVersion === RESOLVED_SECRET_TRACE_CHECKPOINT_VERSION
const valid =
state?.resolvedSecretTraceCheckpointVersion === RESOLVED_SECRET_TRACE_CHECKPOINT_VERSION &&
checkpointPresent &&
state?.sourceExecutionId === options.executionId &&
isResolvedSecretTraceProvenanceV1(provenance) &&
provenance.scope?.workspaceId === options.workspaceId
const registry = new ResolvedSecretTraceRegistry([], valid ? provenance.scope : undefined)
if (!valid) {
registry.markIncomplete()
registry.markIncomplete(
checkpointPresent ? 'backfill-checkpoint-unusable' : 'backfill-checkpoint-absent'
)
return registry
}
await registry.importProvenance(provenance, {
apps/sim/lib/workflows/executor/execution-core.ts
+1 -1
View File
@@ -539,7 +539,7 @@ async function executeWorkflowCoreImpl(
scope: { userId: personalEnvUserId, workspaceId: providedWorkspaceId },
})
if (restoredState && !restoreTrusted) {
resolvedSecretTraceRegistry.markIncomplete()
resolvedSecretTraceRegistry.markIncomplete('restored-provenance-untrusted')
}
if (options.trustedInitialResolvedSecretTraceProvenance !== undefined) {
await resolvedSecretTraceRegistry.importProvenance(
apps/sim/providers/index.test.ts
+1 -1
View File
@@ -1288,7 +1288,7 @@ describe('executeProviderRequest — caller-prepared model input', () => {
it('does not make provider execution depend on registry completeness', async () => {
const incomplete = new ResolvedSecretTraceRegistry()
incomplete.markIncomplete()
incomplete.markIncomplete('unspecified')
await executeProviderRequest(
'anthropic',
apps/sim/providers/runtime-context.test.ts
+4 -4
View File
@@ -527,7 +527,7 @@ describe('provider runtime context', () => {
{ name: 'TOKEN', plaintext: 'secret-value', encryptedValue: 'ciphertext' },
])
mockExecuteTool.mockImplementationOnce(async (_toolId, _params, options) => {
options.resolvedSecretTraceRegistry?.markIncomplete()
options.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
return { success: true, output: { value: 'secret-value' } }
})
@@ -546,7 +546,7 @@ describe('provider runtime context', () => {
{ name: 'TOKEN', plaintext: 'secret-value', encryptedValue: 'ciphertext' },
])
mockExecuteTool.mockImplementationOnce(async (_toolId, _params, options) => {
options.resolvedSecretTraceRegistry?.markIncomplete()
options.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
return {
success: false,
output: { value: 'secret-value' },
@@ -600,7 +600,7 @@ describe('provider runtime context', () => {
{ name: 'TOKEN', plaintext: 'secret-value', encryptedValue: 'ciphertext' },
])
mockExecuteTool.mockImplementationOnce(async (_toolId, _params, options) => {
options.resolvedSecretTraceRegistry?.markIncomplete()
options.resolvedSecretTraceRegistry?.markIncomplete('unspecified')
throw new Error('secret-value')
})
@@ -628,7 +628,7 @@ describe('provider runtime context', () => {
const registry = new ResolvedSecretTraceRegistry([
{ name: 'TOKEN', plaintext: 'secret-value', encryptedValue: 'ciphertext' },
])
registry.markIncomplete()
registry.markIncomplete('unspecified')
mockExecuteTool.mockResolvedValueOnce({
success: true,
output: { value: 'secret-value' },
apps/sim/providers/utils.ts
+1 -1
View File
@@ -1570,7 +1570,7 @@ export function prepareToolExecution(
try {
projectedToolParams = tool.paramsTransform(projectedToolParams)
} catch {
inputRegistry.markIncomplete()
inputRegistry.markIncomplete('tool-params-transform-failed')
projectedToolParams = undefined
}
}
apps/sim/tools/index.test.ts
+18 -2
View File
@@ -1924,7 +1924,7 @@ describe('executeTool Function', () => {
it('runs a private-provenance call from an incomplete parent without replacing its result', async () => {
const registry = new ResolvedSecretTraceRegistry()
registry.markIncomplete()
registry.markIncomplete('unspecified')
const fetchMock = vi.mocked(global.fetch)
const result = await executeTool(
@@ -1938,10 +1938,26 @@ describe('executeTool Function', () => {
expect(fetchMock).toHaveBeenCalled()
})
/**
* The shape that latched in production with `reason: "unspecified"` — a getter or symbol key on
* the params record makes the input lineage unboundable, and the fork is marked before the tool
* runs. Pinned by name so a refusal downstream can be traced back to this guard.
*/
it('names the guard when tool params are not enumerable plain data', async () => {
const registry = new ResolvedSecretTraceRegistry()
const params: Record<string, unknown> = { code: 'return "unreachable"' }
Object.defineProperty(params, 'envVars', { enumerable: true, get: () => ({}) })
await executeTool('function_execute', params, { resolvedSecretTraceRegistry: registry })
expect(registry.isComplete()).toBe(false)
expect(registry.getIncompletenessDiagnostics()?.reasons).toContain('tool-input-not-enumerable')
})
it('runs a private-provenance call when its input lineage cannot be bounded', async () => {
const registry = new ResolvedSecretTraceRegistry()
const incompleteToolRegistry = registry.forkForToolCall()
incompleteToolRegistry.markIncomplete()
incompleteToolRegistry.markIncomplete('unspecified')
vi.spyOn(registry, 'forkForInputPaths').mockReturnValue(incompleteToolRegistry)
const fetchMock = vi.mocked(global.fetch)
apps/sim/tools/index.ts
+1 -1
View File
@@ -1408,7 +1408,7 @@ export async function executeTool(
const toolRegistry = paramEntries
? parentRegistry.forkForInputPaths(paramEntries.map(([key]) => [key] as const))
: parentRegistry.forkForToolCall()
if (!paramEntries) toolRegistry.markIncomplete()
if (!paramEntries) toolRegistry.markIncomplete('tool-input-not-enumerable')
const executionContext = options.executionContext
? { ...options.executionContext, resolvedSecretTraceRegistry: toolRegistry }
: undefined