Do not rebuild account aggregation first
Empower, Kubera, Sharesight, and Snowball already fight this battle. It is integration-heavy and not the clearest wedge.
Start with manual/imported canonical portfolio files and process artifacts.resource map
Competitor analysis for what we should actually build.
Short version: do not build another giant portfolio tracker. Build the missing process layer: canonical portfolio, policy, decision journal, evidence log, review cadence, and an installable agent skill.
Build an Investment Process Score
Simply Wall St and Boldin prove users like quick health snapshots. Portfolio tools score performance; few score decision quality.
Score policy completeness, concentration, drift, evidence freshness, review cadence, and behavior risk.Make agent-maintained local files the core artifact
Obsidian and jsm show that local, installable workflows can be durable and agent-friendly.
Define canonical_portfolio.yaml, investment_policy.md, decision_journal.md, and review_log.md templates.Attach conclusions to external analyses
Portfolio Visualizer/Testfol/Koyfin already do deep analysis. The missing layer is what the investor concluded and committed to.
Let users paste/link analyses, then have the agent extract assumptions, decision, risks, and next review date.Represent portfolios as serving life scenarios
ProjectionLab’s strength is that scenarios feel personal. Portfolio charts alone are abstract.
Every portfolio should answer: what life outcome does this support, and what tradeoff is accepted?highest-value patterns
Things to copy now
Turn each portfolio comparison into a saved thesis: why it fits, what could break it, and when to review.
Create a process-health scorecard: policy fit, concentration, evidence freshness, review discipline, behavior risk.
Represent every portfolio as serving a life scenario, not just maximizing chart metrics.
Create an Investment Process Score, separate from performance: policy completeness, drift, evidence, cadence.
Make local files first-class: canonical_portfolio.yaml, policy.md, decision_journal.md, review_log.md.
Ship an InvestorWorkbench skill that turns a chat agent into a portfolio-process maintainer.
CRM table
Landscape scan
| Tool | Job | Interface | Best feature | Do not reinvent / gap | IW move | Signal |
|---|---|---|---|---|---|---|
| PortfolioChartsAllocation research | Explain lazy portfolios and compare asset-allocation tradeoffs. | Public educational site with embedded charts and long explanations. | Metrics make portfolio recipes feel understandable: CAGR, drawdown, safe withdrawal, ulcer, start-date sensitivity. | Great reference, but not a personal workflow; users leave without a maintained policy or decision record. | Turn each portfolio comparison into a saved thesis: why it fits, what could break it, and when to review. | CopyNow |
| Portfolio VisualizerAllocation research | Run serious backtests, optimization, Monte Carlo, and tactical allocation tests. | Dense forms feeding analytical result pages. | Breadth of analytics modules; it is the spreadsheet power-user's web app. | Analysis is episodic. It does not force a conclusion, decision journal, or process discipline. | Do not rebuild every module first; store conclusions from external analyses and ask the agent to critique assumptions. | AvoidLater |
| Testfol.ioAllocation research | Advanced portfolio experiments: tactical signals, optimized systems, diagnostics. | Power-user research lab with guides and repeatable tests. | Encourages systematic testing rather than casual chart browsing. | Can become too quant-heavy for mainstream DIY investors. | Offer an agent-run robustness checklist instead of exposing every advanced knob in UI. | DifferentiateNext |
| Curvo BacktestAllocation research | Backtest ETF portfolios for European investors. | Localized ETF portfolio builder. | Home-country context and regional fund implementation. | Narrow region/use case. | Add a 'home investor context' field to every canonical portfolio. | CopyNext |
| Snowball AnalyticsPortfolio tracking | Track performance, income, benchmarks, broker imports, and dividends. | SaaS dashboard with broker sync and dividend-forward views. | Dividend calendar plus strategy/category buckets. | Tracking answers what happened, not whether decisions followed a good process. | Map each holding to a strategy bucket and require a written reason for bucket membership. | IntegrateNext |
| SharesightPortfolio tracking | Performance, dividend, and tax reporting for real portfolios. | Tracker organized around holdings, reports, brokers, and tax periods. | Performance/tax reporting is concrete and high-value. | Not a research or policy system. | Export a process report that sits beside tax/performance reports: decisions, rationales, policy drift. | IntegrateLater |
| KuberaPortfolio tracking | Track complex net worth across public, private, real, crypto, and entities. | Premium live balance sheet. | Asset breadth and nested entities. | Too broad/expensive for a first DIY portfolio-process wedge. | Use a simple YAML asset model first; add account sync much later. | AvoidLater |
| EmpowerPortfolio tracking | Free aggregation, budget, net worth, portfolio mix, retirement overview. | Consumer dashboard that can become an advisor funnel. | Complete financial picture with low adoption friction. | User is pushed toward advisory/managed relationship; DIY process is not the product. | Stay explicitly DIY and user-owned; no hidden advisor funnel. | DifferentiateLater |
| Simply Wall StPortfolio tracking | Make stock/portfolio research visual and approachable. | Visual command center with snowflake graphics, alerts, narratives, and community. | One-glance visual storytelling. | Pretty visuals can oversimplify and encourage stock picking attention loops. | Create a process-health scorecard: policy fit, concentration, evidence freshness, review discipline, behavior risk. | CopyNow |
| KoyfinMarket research | Market, macro, ETF, stock, news, and advisor-grade dashboards. | Configurable financial terminal. | Macro and market dashboard depth. | Research terminal can become endless information consumption. | Translate macro indicators into portfolio review prompts, not trading signals. | IntegrateLater |
| TradingViewMarket research | Charts, indicators, scripts, calendar, community ideas, broker execution. | Charting-first social platform. | Scriptable indicators and community idea marketplace. | Optimized for trading attention, not long-term process discipline. | Create installable review checklists, not trading indicators. | DifferentiateLater |
| Stock RoverMarket research | Screen stocks, analyze portfolios, valuations, rebalancing, and market briefs. | Dense independent-investor research cockpit. | Screeners and portfolio analytics in one place. | Can reward constant tinkering. | Use screens only as evidence attached to decisions with timestamped snapshots. | IntegrateLater |
| Portfolio123Market research | Build systematic stock strategies with ranking, simulations, and AI/ML. | Strategy lab: idea -> rank -> backtest -> launch. | Explicit research pipeline for systematic rules. | Overkill unless user is building active strategies. | Apply the same pipeline to portfolio policy changes: hypothesis -> evidence -> decision -> review date. | CopyNext |
| ProjectionLabFinancial planning | Model financial independence, retirement, taxes, estate, and life scenarios. | Beautiful scenario simulator / living model. | Scenario planning feels personal and visual; strong privacy choices. | Life-plan model is not the same as investment decision governance. | Represent every portfolio as serving a life scenario, not just maximizing chart metrics. | CopyNow |
| BoldinFinancial planning | Retirement readiness, what-if planning, tax strategies, wellness score, coach support. | Guided planner with wellness score and scenario comparisons. | Score + what-if framing gives users confidence quickly. | Planner scope is broad; investment process is a subset. | Create an Investment Process Score, separate from performance: policy completeness, drift, evidence, cadence. | CopyNow |
| ObsidianProcess / agent layer | Own local markdown notes, templates, backlinks, journals, and periodic reviews. | File-first personal knowledge base. | User-owned files and flexible note workflows. | No opinionated investment process out of the box. | Make local files first-class: canonical_portfolio.yaml, policy.md, decision_journal.md, review_log.md. | CopyNow |
| jsm / Skills.mdProcess / agent layer | Install, sync, version, inspect, and run agent skills as operational workflows. | Public skill catalog plus CLI with local cache and JSON output. | Makes agent workflows portable and repeatable. | Generic skill infrastructure, not investor-specific. | Ship an InvestorWorkbench skill that turns a chat agent into a portfolio-process maintainer. | CopyNow |
positioning
The wedge
Existing products are good at analysis, tracking, research, or planning. The underserved layer is investment governance for DIY investors: why the portfolio exists, what evidence supports it, what would cause a change, and whether the user is following their own process.
agent onboarding
The skill
Inspired by jsm: ship an installable InvestorWorkbench skill. The user chats with Claude/Codex/Hermes; the agent maintains local files; the web app reads them into dashboards and reports.
- Install skill
- Create canonical_portfolio.yaml + policy.md
- Capture decisions into decision_journal.md
- Run scheduled review/checkup prompts
Allocation research4
Portfolio tracking5
Market research4
Financial planning2
Process / agent layer2