Skip to content
Original file line number Diff line number Diff line change
Expand Up @@ -203,6 +203,17 @@ Workflows can also wait for multiple external event signals of the same name, in

Learn more about [external system interaction.]({{% ref "workflow-patterns.md#external-system-interaction" %}})

## Workflow history propagation

A parent workflow can opt to share its execution history with downstream child workflows and activities. Two scopes are available:
Comment thread
cicoyle marked this conversation as resolved.

- **Lineage** — caller's events plus the full ancestor chain (chain-of-custody)
- **Own history** — caller's events only (a trust boundary, ancestral lineage dropped)

The receiving workflow/activity reads its inherited history via `ctx.GetPropagatedHistory()` and can verify what happened upstream — useful for fraud checks, compliance gates, and long-running AI agents that need context across hops.
Comment thread
cicoyle marked this conversation as resolved.
Outdated

For details, the scope comparison, and code examples, see [History propagation]({{< ref workflow-history-propagation.md >}}).
Comment thread
cicoyle marked this conversation as resolved.
Outdated

## Purging

Workflow state can be purged from a state store, purging all its history and removing all metadata related to a specific workflow instance. The purge capability is used for workflows that have run to a `COMPLETED`, `FAILED`, or `TERMINATED` state.
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,138 @@
---
type: docs
title: "Workflow history propagation"
linkTitle: "History propagation"
weight: 2500
description: "Share a parent workflow's execution history with child workflows and activities for chain-of-custody, audit, and AI-agent context"
---

By default, a child workflow or activity only sees the input it was scheduled with. The parent's execution history — what activities ran, what child workflows it spawned, what events it observed — is invisible. Workflow history propagation lets a parent opt-in to share that history with the work it schedules.
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would add a some scenarios into this article as well, in case most likely, that some comes straight to this article from a search (like I suggest in the overview)

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would add a some scenarios into this article as well, in case most likely, that some comes straight to this article from a search (like I suggest in the overview)

Comment thread
cicoyle marked this conversation as resolved.
Outdated

This is useful for:

- **Chain-of-custody and audit** — a fraud check, compliance gate, or settlement step can verify what already happened upstream rather than trusting the caller's claims.
- **Cross-app workflows** — a downstream app, often owned by a different team, can confirm an upstream activity ran without an out-of-band query.
Comment thread
cicoyle marked this conversation as resolved.
- **AI agents and long-running orchestrations** — agents that loop with `ContinueAsNew` or fan out to sub-agents can carry context (tool calls, model outputs, intermediate decisions) forward without re-prompting from scratch.

## Propagation scopes

Two scopes are exposed. The default is no propagation.

| Scope | What gets sent | When to use it |
| ----- |-----------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------|
| `PropagateLineage` | Caller's own events plus the full ancestor chain inherited from the caller's own parent | Full chain-of-custody, downstream wants to see everything that happened before, all the way to the root |
| `PropagateOwnHistory` | Caller's own events only, ancestral lineage is dropped | Trust boundary — the caller is willing to vouch for what *it* did but the receiver shouldn't see further upstream |
Comment thread
cicoyle marked this conversation as resolved.
Outdated

`PropagateOwnHistory` is the trust boundary: choosing it tells the runtime to stop forwarding any history the caller itself received. This is the right choice when the receiver is a less-trusted app, a third party, or operates under different compliance rules.

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Any more guidance we can provide here? For example PropagateOwnHistory only has to be used if you have no child workflows, but this is very unlikely typically.
Should you typically always use Lineage in 90% of cases?
How can you set this on a Dapr agent (also an Langgraph agent that has been made durable with workflows)

## Setting it up

A parent opts a single `CallActivity` or `CallChildWorkflow` into propagation via the `WithHistoryPropagation` option. Other activity / child calls in the same workflow are unaffected.

```go
Comment thread
cicoyle marked this conversation as resolved.
import (
"github.com/dapr/durabletask-go/api"
"github.com/dapr/durabletask-go/task"
Comment thread
cicoyle marked this conversation as resolved.
Outdated
)

func ParentWorkflow(ctx *workflow.WorkflowContext) (any, error) {
// Activity does NOT receive propagated history (default).
if err := ctx.CallActivity("validateMerchant").Await(nil); err != nil {
return nil, err
}

// Child workflow DOES receive parent's history (LINEAGE).
var result string
if err := ctx.CallChildWorkflow("processPayment",
task.WithHistoryPropagation(api.PropagateLineage()),
).Await(&result); err != nil {
return nil, err
}
return result, nil
}
```

## Receiving propagated history
Comment thread
cicoyle marked this conversation as resolved.
Outdated

Inside a child workflow or activity, call `ctx.GetPropagatedHistory()`. It returns the propagated history if the caller opted in, or `nil` if it didn't.

```go
func FraudCheck(ctx *workflow.WorkflowContext) (any, error) {
ph := ctx.GetPropagatedHistory()
if ph == nil {
return "no upstream history", nil
}

// The history exposes both raw events and per-app/per-instance chunks.
fmt.Printf("scope: %s, %d events from apps %v\n",
ph.Scope(), len(ph.Events()), ph.GetAppIDs())

// Drill into a specific upstream workflow's activities:
merchantWf, err := ph.GetWorkflowByName("MerchantCheckout")
if err != nil {
return nil, fmt.Errorf("expected MerchantCheckout in propagated history: %w", err)
}
Comment thread
cicoyle marked this conversation as resolved.
validation, err := merchantWf.GetActivityByName("ValidateMerchant")
if err != nil {
return nil, fmt.Errorf("expected ValidateMerchant in propagated history: %w", err)
Comment thread
cicoyle marked this conversation as resolved.
Outdated
}
if !validation.Completed {
return "merchant validation didn't complete, rejecting", nil
}

return "approved", nil
}
```

The returned `PropagatedHistory` carries:

- `Events()` — the flat list of upstream `HistoryEvent`s in order
- `Scope()` — which scope the parent chose (`OWN_HISTORY` or `LINEAGE`)
- `GetWorkflows()` — per-workflow chunks (each is a `WorkflowResult` tagged with `appId`, `workflowName`, `instanceId`, `startEventIndex`, `eventCount`)
- `GetAppIDs()` — deduplicated list of every app that contributed events
- Top-level filters: `GetWorkflowByName(name)`, `GetWorkflowsByName(name)`, `GetEventsByAppID(appID)`, `GetEventsByInstanceID(id)`, `GetEventsByWorkflowName(name)`

Each `WorkflowResult` further exposes:

- `GetActivityByName(name)` / `GetActivitiesByName(name)` — find activities the upstream workflow ran. Returns an `ActivityResult` with `Completed bool` and the result payload.
- `GetChildWorkflowByName(name)` / `GetChildWorkflowsByName(name)` — find child workflows the upstream workflow scheduled.

The singular `*ByName` methods return `ErrPropagationNotFound` when no match exists, the plural variants return an empty slice.

## Cross-app workflows
Comment thread
cicoyle marked this conversation as resolved.
Outdated

When a parent workflow in App A calls a child workflow in App B, the propagated chunks travel between sidecars. Two security knobs apply:
Comment thread
cicoyle marked this conversation as resolved.
Outdated

- **mTLS** — when Dapr is deployed with mTLS (the default for Helm / `dapr init -k`), inter-sidecar traffic is encrypted and authenticated.
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What does mTLS have to do with propagation? The statement below just say communication is encrypted.

- **`WorkflowHistorySigning`** — an opt-in `Configuration` feature that signs each propagated chunk with the producing app's SPIFFE identity. Receivers can verify chunks weren't tampered with after they left the producer.
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Make this its own heading and then reference the article in this PR
#5082
`
``suggestion

Workflow history signing.

With history signing, the propagated context can be cryptographically verified. For more information see [workflow history signing]({{% ref "workflow-history-signing.md" %}}


If `WorkflowHistorySigning` is not enabled, daprd logs a warning per dispatch:

> `propagating unsigned workflow history to child workflow '...' (signing is not configured; chunks cannot be cryptographically verified by the receiver)`
Comment thread
cicoyle marked this conversation as resolved.
Outdated

{{% alert title="Treat unsigned chunks as untrusted" color="warning" %}}
Without signing, propagated chunks are functional but not cryptographically verifiable. Don't treat unsigned propagated history as authoritative for high-value decisions (payments, approvals). Enable `WorkflowHistorySigning` for production deployments that depend on chain-of-custody.
Comment thread
cicoyle marked this conversation as resolved.
Outdated
{{% /alert %}}

## ContinueAsNew and rerun

Propagated history flows correctly through both `ContinueAsNew` and rerun:
Comment thread
cicoyle marked this conversation as resolved.
Outdated

- A workflow that received propagated history and calls `ContinueAsNew` passes the same incoming chunks to its next generation.
Comment thread
cicoyle marked this conversation as resolved.
Outdated
- Rerunning a workflow re-issues activity / child-workflow calls with the same propagation scope they were originally scheduled with — `LINEAGE` stays `LINEAGE`, `OWN_HISTORY` stays `OWN_HISTORY`.
Comment thread
cicoyle marked this conversation as resolved.
Outdated

This makes long-running agents and crash-recovery scenarios behave the way you'd expect: the receiving generation/rerun sees the same history the original run did.

## Next steps

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Discover [how to apply workflow signing]({{% ref "workflow-history-signing.md" %}} to attest the validity of a previous workflow or activity step.

{{< button text="Workflow patterns >>" page="workflow-patterns.md" >}}

## Related links

- [Workflow overview]({{< ref workflow-overview.md >}})
- [Workflow features and concepts]({{< ref workflow-features-concepts.md >}})
- [Workflow architecture]({{< ref workflow-architecture.md >}})
- [How to author a workflow]({{< ref howto-author-workflow.md >}})
- [Workflow API reference]({{< ref workflow_api.md >}})
- Try out the full SDK example:
- [Go example](https://github.com/dapr/go-sdk/tree/main/examples/workflow-history-propagation)
Original file line number Diff line number Diff line change
Expand Up @@ -58,6 +58,12 @@ Same as Dapr actors, you can schedule reminder-like durable delays for any time

[Learn more about workflow timers]({{% ref "workflow-features-concepts.md#durable-timers" %}}) and [reminders]({{% ref "workflow-architecture.md#reminder-usage-and-execution-guarantees" %}})

### History propagation

A parent workflow can opt to share its execution history with child workflows and activities — useful for chain-of-custody verification, fraud detection, audit, and AI-agent context that must flow across hops.
Comment thread
cicoyle marked this conversation as resolved.

[Learn more about workflow history propagation.]({{< ref workflow-history-propagation.md >}})

### Workflow HTTP calls to manage a workflow

When you create an application with workflow code and run it with Dapr, you can call specific workflows that reside in the application.
Expand Down
Loading