Skip to content

Getting Started

This guide walks through setting up your investment data — theses, watchlist, and running your first workflows. For installation and API key setup, see the home page.

Soul & Policy

During openfin init, you configure two files that shape how the AI agent reasons about your portfolio:

  • Soul (~/.openfin/soul.yaml) — a prose narrative describing who you are as an investor. Are you concentrated or diversified? Tax-sensitive? Comfortable with drawdowns? The agent reads this verbatim and adapts its analysis to your style.
  • Policy (~/.openfin/policy.yaml) — guidance and hard limits. Pick from three starter templates during init, then customize. Limits (max position %, max sector %, min cash %) are enforced by automation; guidance prose shapes the agent's recommendations.
openfin policy show              # view your active soul + policy
openfin shell                    # start a Claude Code session with soul/policy/CLI context preloaded

You can edit either file at any time — they're plain YAML. The agent picks up changes on the next run.

Investment Theses

Theses are the core of the system. A thesis captures an investment story — not just a ticker — and maps it to the symbols it implicates. Everything downstream (watchlist, daily brief, weekly review) derives from your theses.

Creating a Thesis

Each thesis is a YAML file at ~/.openfin/theses/{slug}.yaml (seeded from repo examples by openfin init). Here's an example:

slug: ai-compute-hardware
title: AI Drives Compute Hardware Demand
narrative: >
  Hyperscaler and enterprise AI training/inference workloads continue to
  grow, driving sustained demand for GPUs, custom silicon, and networking
  chips. The capex cycle extends as models get larger and inference scales out.
status: active                      # active | archived | draft
symbols:
  - ticker: NVDA
    role: primary                   # primary | secondary | hedge | adhoc
    conviction: starter             # starter | half | full
    conditions:
      add:
        - metric: pe_ratio
          operator: "<"
          value: 30
          description: "Valuation pullback entry"
      trim:
        - metric: pe_ratio
          operator: ">"
          value: 80
    targets:
      price_target: 200
      stop_loss: 100
  - ticker: TSM
    role: secondary
    conviction: starter
    conditions:
      add:
        - metric: pe_ratio
          operator: "<"
          value: 20
          description: "Foundry bottleneck — all AI silicon depends on TSMC"
    targets:
      price_target: null
      stop_loss: null
catalysts:
  - date: "2026-05-28"
    description: "NVDA Q1 FY27 earnings — data center revenue trajectory"
qualitative:
  confirming:
    - "Data center revenue growth sustaining QoQ"
    - "Hyperscaler capex guides increasing or holding"
  disconfirming:
    - "Hyperscaler capex pullback or deferrals"
    - "In-house ASICs displacing merchant GPU at meaningful scale"
notes: "Core AI hardware thesis"
created: "2026-03-18"

A thesis starts with a narrative, not a ticker. Each symbol gets a role, conviction level, and optional entry/trim/exit conditions. See the Thesis Management SOP for the full creation workflow.

Symbol Roles

Role Meaning Example
primary Direct, high-conviction exposure to the thesis NVDA in AI compute
secondary Related or supporting exposure TSM as foundry for AI chips
hedge Benefits if the thesis partially fails Airlines bought on oil-fear dip
adhoc Watching, no strong thesis tie yet One-off interesting stock

Verifying Theses

openfin thesis list              # all theses with status and symbols
openfin thesis show ai-compute   # full details for a thesis
openfin thesis status ai-compute # health, time pressure, active hypotheses
openfin thesis symbols           # show all symbols and which theses they belong to

Watchlist

The watchlist is derived automatically from active theses and portfolio holdings. There is no manual watchlist file to maintain.

openfin watchlist list            # show derived watchlist with live prices
openfin watchlist sources         # show where each symbol comes from (thesis slugs, portfolio)

For one-off stocks not tied to a specific story, use the adhoc watchlist:

openfin watchlist add FOO         # adds to the special "watchlist" thesis
openfin watchlist remove FOO      # removes from it

First Workflows

Once your theses are set up, try these workflows.

Check Your Portfolio

openfin brokerage accounts              # connected accounts and balances
openfin brokerage positions --all       # current holdings with gain/loss

Research a Symbol

openfin market quote NVDA                       # live quote with 52w range, P/E
openfin research news NVDA                      # recent headlines
openfin research sec-filings NVDA --type 10-K   # SEC filings
openfin research search "AI inference demand"   # investment-focused web search
openfin thesis status ai-compute                 # thesis health + active hypotheses

Market Overview

openfin market overview --period 5d     # index and sector performance
openfin market macro                    # FRED economic indicators
openfin market earnings                 # upcoming earnings for watchlist

Run a Daily Brief

The daily brief gathers pre-market intelligence: earnings alerts, thesis-relevant news, catalyst alerts, and market context.

openfin review daily                    # generate today's brief
openfin review daily --date 2026-03-18  # specific date
openfin review daily --dry-run          # preview without persisting

When running via the Telegram bot with --ai, the brief includes AI-generated analysis with actionable takeaways.

Run a Weekly Review

The weekly review collects portfolio, market, macro, thesis, and filing data into a structured report, then optionally runs an AI agent to score symbols and produce decisions.

openfin review weekly --dry-run         # preview without writing artifacts
openfin review weekly                   # full data collection + report
openfin review weekly --ai              # data collection + headless agent analysis

See the CLI Reference for the full review pipeline including scoring, evidence, annotation, and finalization commands.

Next Steps