Dossier
    Automated Project Tracking Dashboard (MLNNA)Updated Jun 13, 2026

    Project Intake - Automated Project Tracking Dashboard (MLNNA)

    Markdown file

    Imported markdown

    Project Intake - Automated Project Tracking Dashboard (MLNNA)

    1) Executive Summary

    Automated Project Tracking Dashboard (MLNNA) is best represented as the Notion-backed portfolio operations dashboard implemented in notion-project-dashboard-automation, with earlier planning lineage in AgentsOS-Project-Dashboard and a later Node/Express rewrite in Dashboard-Automator.

    Current state should be imported conservatively. The latest source commit is deea8e94943ebdcd68405d9e6eff79e175ad1417 with subject chore: disable repo automations on 2026-03-22. Treat the project as paused/reference or operator-only unless a current production dashboard deployment is separately verified.

    2) Project Type and Domain

    • Project type: experiment.
    • Type: internal portfolio operations dashboard and automation system.
    • Domains: builder, entrepreneur.
    • Trail: AI Ops.
    • Status: paused/reference for automation current-state claims, with public MLNNA surfaces preserved as existing project links.
    • Domain: AI Ops, repo portfolio governance, project tracking, Notion-backed operating system.
    • Primary user: single builder/operator managing many GitHub repositories and AI-assisted workstreams.
    • Secondary future user: small teams managing shared initiative portfolios, documented as future/draft in product docs.
    • Project posture: source-backed implementation with public dashboard mode, but current production operation is unconfirmed.

    3) Repository and Live URLs

    4) Product/User Journey Summary

    The core journey is a recurring project-review loop:

    1. The operator configures tracked repositories and Notion access.
    2. The sync pipeline pulls GitHub repository metadata and recent activity into Notion.
    3. The operator logs manual conversations, tasks, and decision context against repositories.
    4. AI synthesis creates per-repo goals, portfolio phase/purpose fields, focus warnings, and recommendations.
    5. The snapshot exporter writes JSON/CSV data for dashboard rendering.
    6. The dashboard shows portfolio KPIs, project phase distribution, at-risk projects, project health cards, goals, weekly activity, and repo-level drilldowns.
    7. The operator reviews the outputs, adjusts Notion/project context, and reruns the loop.

    The dashboard has both private and public modes. The public mode presents a landing page at / and read-only dashboard at /dashboard; private mode expects authentication and private snapshot access. Existing Hyphenomenon catalog context also points to a public MLNNA site and Lovable-hosted prototype, so the imported project page should show those as project surfaces without treating them as proof that the paused automation is currently running.

    5) Architecture and Stack Summary

    • Data backbone: Notion databases for Projects, GitHub Activity, Conversations, Tasks, Goals, and AI Runs.
    • Dashboard app: React 18, Vite, TypeScript, Tailwind CSS 4, shadcn-style UI primitives, Recharts, lucide icons.
    • Auth/deployment: Better Auth and Vercel dashboard app wiring; docs site uses VitePress and Cloudflare Pages.
    • Runtime data: private snapshot under apps/dashboard/data/private/snapshot.json; public-safe snapshot under apps/dashboard/public/data/snapshot.public.json.
    • Automation: repo docs describe GitHub Actions scheduled sync/export; current repo state also includes a disabled automation marker and latest commit disabling repo automations.
    • Tests: Python unittest suite for Notion bootstrap shape, GitHub to Notion sync, snapshot export, AI synthesis, publishing, config, conversation logging, and Codex Wrapped integration.

    6) Key Workflows and Operational Flows

    • Notion bootstrap: idempotently creates/updates dashboard databases under the configured Notion parent page.
    • GitHub sync: reads allowlisted repositories, pulls recent GitHub activity, and upserts project/activity records by stable keys.
    • Conversation logging: CLI commands create conversation records and optionally task records linked to a repo/project.
    • Repo-level AI synthesis: builds evidence bundles and writes structured goals with status, priority, confidence, rationale, and evidence links.
    • Portfolio AI synthesis: infers purpose, phase, top goal, risk flag, focus score, and recommendations across the full repo portfolio.
    • Snapshot export and publish: exports normalized JSON/CSV and copies dashboard data into private and public app data paths.
    • Dashboard review: operator checks KPIs, project health, goals, activity, and project detail timelines from the React dashboard.
    • Human review boundary: AI outputs are decision support and Notion remains editable when inference is imperfect.

    7) AI Context: Prompts, Chats, Agent Workflows

    The source repo includes AI synthesis modules rather than a raw chat archive. The AI context is operational:

    • src/dashboard/ai_synthesis.py performs repo-level goal synthesis from recent activity, manual logs, repo metadata, and existing context.
    • src/dashboard/ai_portfolio.py performs cross-project analysis and writes strategic fields back to Notion Projects.
    • DOCS/strategist-logic.md documents the strategy model: data layer, interpretation layer, portfolio layer, visibility layer, and action layer.
    • The project uses schema-constrained AI outputs and audit rows in AI Runs, but the intake did not review raw prompts or provider logs beyond source files and docs.
    • Related agent runtime lineage exists in AgentsOS-Project-Dashboard, which planned Notion -> Observable snapshot flows before the React dashboard implementation.

    8) Supporting Docs and External Resources

    Hydration Source Artifacts

    Artifact (relative repo path)Canonical URLTypeWhy hydrate this into note body
    README.mdhttps://github.com/maggielerman/notion-project-dashboard-automation/blob/main/README.mdmarkdownPrimary command surface and repository description.
    DOCS/solution-implementation.mdhttps://github.com/maggielerman/notion-project-dashboard-automation/blob/main/DOCS/solution-implementation.mdmarkdownBest source for architecture, pipeline, and runtime flow.
    DOCS/strategist-logic.mdhttps://github.com/maggielerman/notion-project-dashboard-automation/blob/main/DOCS/strategist-logic.mdmarkdownBest source for the project decision model and AI strategist framing.
    apps/dashboard/src/App.tsxhttps://github.com/maggielerman/notion-project-dashboard-automation/blob/main/apps/dashboard/src/App.tsxtextSource proof for public/private modes, dashboard tabs, and user journey.
    ./screenshots/02-primary-workflow-overview.pngUnknownimageCaptured proof of the public dashboard overview rendered locally.
    UnknownUnknownpdfNo .pdf source artifact was found in the reviewed source repo.

    9) Screenshot Gallery (with relative links + captions)

    1. Primary Workflow: Dashboard overview
      Read-only dashboard overview with portfolio KPIs, project phases, risk list, health cards, and activity context.
    2. Key Feature: Project health grid
      Projects tab showing repo filters and health cards sorted by health score.
    3. Admin/Settings Equivalent: Project drilldown
      MLNNA source repo drilldown with activity, CI failures, passes, issues, and repo-specific activity chart.
    4. Workflow View: Goals board
      Goals board grouped across Now, Next, Later, and Done states.
    5. Activity/Monitoring View: Activity feed
      Activity view with weekly activity visualization and recent repository events.

    10) Candidate Import Nodes and Relationships

    • Project supernode:
      • Title: Automated Project Tracking Dashboard (MLNNA)
      • Candidate slug: mlnna-dashboard-automation
      • Candidate status: paused or reference until current production operation is confirmed.
      • Trail: AI Ops
    • Source repo node:
      • notion-project-dashboard-automation
      • Relationship: implements primary MLNNA pipeline and dashboard.
    • Workflow node:
      • mlnna-portfolio-sync-and-review-loop
      • Relationship: describes GitHub/Notion sync, AI synthesis, snapshot export, dashboard review, and human correction.
    • Supporting repo nodes:
      • AgentsOS-Project-Dashboard: planning lineage for Notion plus Observable snapshots.
      • Dashboard-Automator: later Node/Express rewrite/successor with authenticated app surfaces, Neon, sync history, and AI run evaluation.
    • Data model candidates:
      • Notion portfolio operations schema: Projects, GitHub Activity, Conversations, Tasks, Goals, AI Runs.
      • Snapshot/public-dashboard data contract: projects, goals, activity, conversations, tasks, portfolio summary.
    • Evidence relationships:
      • README -> project summary.
      • Solution implementation -> architecture and workflow.
      • Strategist logic -> AI/context interpretation layer.
      • Screenshots -> rendered public dashboard proof.

    11) Risks, Gaps, and Unknowns

    • Confirmed production dashboard URL is unknown. The verified dashboard screenshots came from local public mode at route /dashboard on port 3071.
    • The previous Hyphenomenon static project record linked https://mlnna.com and https://mlnna.lovable.app; those links are preserved here, but this intake did not verify their deployment freshness or relationship to the paused source automation.
    • The source repo's latest commit disables repo automations, so the project should not be represented as currently self-running without fresh operational proof.
    • MLNNA is not consistently expanded in source docs. It appears as an operator/project label, while source infrastructure includes mlnna-dashboard and mlnna.pages.dev.
    • No PDF or DOCX source artifact was found in the primary source repo during this intake.
    • The public snapshot data is useful for route proof but should not be overtreated as live-current telemetry without checking deploy/runtime freshness.
    • Dashboard-Automator overlaps with this problem space but is not the same source repo; it should be related lineage or a separate project/update decision, not silently merged into the primary source record.
    • No local or production knowledge-backbone apply was run in this intake.

    12) Handoff Checklist for Hyphenomenon Import

    • Confirm dossier sections complete and source-backed.
    • Confirm screenshot files exist in DOCS/intake/runs/projects/2026-06-13-mlnna-project-tracking-dashboard/screenshots.
    • Confirm screenshot links render from DOCS/intake/runs/projects/2026-06-13-mlnna-project-tracking-dashboard/dossier.md.
    • Confirm Unknowns are explicit and no production dashboard operation is fabricated.
    • Create project-page-packet.json for project-page narrative, product surfaces, tech stack, key dates, data model highlights, and data model artifact.
    • Create workflow-packet.json for the recurring sync/synthesis/review loop.
    • Create source-artifact-hydration.json for curated hydration candidates.
    • Run project intake validation with hydration and project-page packet.
    • Run project sync dry-run and review output.
    • Run workflow packet dry-run and review output.
    • Apply locally after operator approval.
    • Apply to production only after separate operator approval and route/provenance verification.

    Workflow packet dry-run note: attempted on 2026-06-13 09:38 EDT (America/New_York), but the script requires DATABASE_URL / POSTGRES_URL / NEON_DATABASE_URL even without --apply. The failure output is preserved in workflow-packet-dry-run.txt.

    Merge correction checkpoint: 2026-06-13 10:38 EDT (America/New_York). Operator review showed the first local overwrite did not carry forward the existing MLNNA public surfaces. The dossier and project-page packet were updated to preserve the existing Hyphenomenon catalog links https://mlnna.com and https://mlnna.lovable.app, while keeping the fresh source-backed automation record paused until current production operation is verified. Corrected dry-run project-sync-dry-run-local-db-merged.txt planned project type experiment, status paused, 3 website links, 3 narrative sections, 6 product surfaces, and no guard failures. Corrected apply run project-intake-apply-2026-06-13T14:38:06.647Z reapplied local row 4012 with those public links.

    Operator trim checkpoint: 2026-06-13 10:43 EDT (America/New_York). Removed the landing-page screenshot from the dossier gallery and artifact folder, removed the extra preservation narrative section and matching key-date entry from the project-page packet, and reapplied local row 4012 with run project-intake-apply-2026-06-13T14:42:35.995Z. Completed since prior checkpoint: validation passed, dry-run project-sync-dry-run-local-db-trimmed.txt planned 5 screenshots and 2 narrative sections, apply uploaded 5 media objects with 0 media failures, and local route HTML verification confirmed the trimmed content is absent while the dashboard overview remains. Next before the next checkpoint: operator reviews the refreshed local page and decides whether more trimming is needed before any separate production dry-run/apply.

    Dataset Preview

    • Raw CSV row/table content is available in the source artifact.