See every step.
Track every retry.
Records every step execution event — attempts, retries, durations, errors, dependencies — with timestamped events and export to JSON, NDJSON, diagrams, and an interactive HTML dashboard.
package main
import (
"context"
flow "github.com/Azure/go-workflow"
auditlog "github.com/larsartmann/go-workflow-auditlog"
)
func main() {
audit, _ := auditlog.New(auditlog.Config{
Enabled: true,
WorkflowID: "data-pipeline",
})
w := &flow.Workflow{}
// add your steps...
audit.Attach(w) // 1. Inject callbacks
_ = w.Do(context.Background()) // 2. Run workflow
audit.Snapshot(w) // 3. Capture final state
_ = audit.ExportHTML("report.html") // Interactive dashboard
}Full visibility into every workflow run.
Every decision optimized for observability — from per-attempt events to interactive dashboards.
Per-Attempt Events
Every retry, every duration, every error — captured as timestamped events with attempt numbers and phases.
Full DAG Structure
Dependency graph, retry/timeout config, step types, skipped and canceled statuses — all captured automatically.
Interactive Dashboard
Self-contained HTML report with DAG graph engine, sortable tables, Gantt timeline, and collapsible tree — zero dependencies.
12+ Export Formats
JSON, NDJSON, Mermaid, PlantUML, Graphviz DOT, D2, tables (16 sub-formats), ASCII tree, HTML tree, and interactive HTML.
Report Diffing
Compare two runs for regression detection — added, removed, and changed steps with duration deltas and status changes.
Cross-System Correlation
128-bit crypto-random RunID stamped on every event for trace/log correlation across distributed systems.
Three calls. Full audit trail.
Attach, run, snapshot. The library handles the rest — callbacks fire per-attempt, capturing everything automatically.
Attach
Inject BeforeStep / AfterStep callbacks into every step via State.MergeConfig. Call before Do.
audit.Attach(w) Execute
During w.Do(ctx), callbacks fire per-attempt, recording timestamped events with duration and error.
w.Do(ctx) Snapshot
Read post-execution state to capture DAG structure, retry config, and skipped/canceled statuses.
audit.Snapshot(w) Why go-workflow-auditlog?
Manual logging is high-boilerplate and easy to forget. This library captures everything automatically.
| No audit | Manual logging | go-workflow-auditlog | |
|---|---|---|---|
| Per-attempt events | ~ | ||
| DAG structure | |||
| Skipped/canceled detection | |||
| Retry & timeout config | |||
| HTML dashboard | |||
| Diagram export | |||
| Report diffing | |||
| RunID correlation | |||
| Lines of boilerplate | 0 | ~50 | 3 |
Built for real workflows.
Drop it into any go-workflow pipeline for instant observability.
CI/CD Pipelines
Audit every step execution, retry, and failure for post-mortem analysis
Data Pipelines
Track transform chain timing for bottleneck detection and regression alerts
Microservice Orchestration
Correlate workflow runs with distributed traces via 128-bit RunID
Compliance & Audit Trails
Immutable event stream for regulated workflows and SOC2/HIPAA evidence
Start auditing in minutes.
Three calls. Zero boilerplate. Full visibility.