MLNNA portfolio sync and review loop
The workflow that turns scattered repository activity and manual context into a reviewable portfolio dashboard: sync GitHub activity into Notion, attach conversation/task context, run AI synthesis, export snapshots, review rendered dashboard evidence, and let the human operator correct the source-of-truth records before the next run.
Workflow steps
- 1
Operator decides whether to correct Notion/source context, accept the current evidence, or rerun the workflow before publication
Decision
- 2
Operator configures Notion access, GitHub token, repository allowlist, and dashboard environment before running the loop
Action
- 3
GitHub sync reads configured repositories and upserts project and activity records into Notion with stable keys
Process
- 4
Operator logs conversations, tasks, and decision context so source records include more than raw repo telemetry
Action
- 5
AI repo synthesis builds evidence bundles and writes structured goals with confidence, rationale, and evidence links
Process
- 6
AI portfolio analysis infers purpose, phase, top goal, risk flag, focus signals, and recommendations across all projects
Process
- 7
Snapshot export creates JSON and CSV artifacts that feed private and public dashboard data paths
Action
- 8
Human operator reviews dashboard proof surfaces including KPIs, project health, goals, activity, and repo drilldowns
Decision