Fill in a task table
Pricing options
Pay per useCN¥0.20 / task
Instructions and workflow details
What to prepare
Research each U.S. stock using a value-investing framework: business model, competitive moat and capital allocation, financial quality, valuation, and margin of safety. The cloud workflow prepares three-statement analysis, a DCF, an evidence audit, and a report. For research assistance only; not investment advice. It provides no price targets, return forecasts, or buy/sell recommendations.
- As-of date (optional)Optional
- YYYY-MM-DD used for freshness checks. Blank uses modelAssumptions.asOf or the run date.
- Ticker data packet (JSON string)Required
- Three JSON-encoded packets concatenated as one string: identityPacket, financialPacket, qualitativePacket. The local Skill's preflight validates freshness, HTTPS sources, and unit consistency before this template is invoked.
- Model assumptions (JSON string)Required
- Explicit DCF drivers for bear / base / bull: years, revenue growth, EBIT margin, tax rate, D&A, capex, NWC, interest, debt repayment, payout, buyback, WACC, terminal method and value. Missing drivers block the model; there are no automatic defaults.
- Override model (optional; leave blank for default)Optional
- Pick a specific model from this Listing's eligible set, or leave blank to use the server's current default.
Generation instructions
Each row is submitted as a separate task. Required fields are marked with *.
- One US equity per row. Supply tickerPacket (the three packets plus balanceSheet and marketCapUsd) and modelAssumptions; optionally supply asOf for freshness checks.
- Step 0 is a deterministic Code node that validates the input contract. If ready=false, all downstream steps are skipped and no model or API fees are incurred for them.
- Four parallel reviews run per row when preflight passes: business model & circle of competence, moat & management, financial quality, valuation with margin of safety.
- The model step runs deterministic Python code: integrated three-statement projection, three-scenario DCF, reverse DCF, WACC x terminal sensitivity, and an xlsx workbook artifact. Missing drivers block the model.
- The audit step resolves evidence paths, checks narrative numbers, scores the four dimensions, and validates model self-checks. The report step always runs and labels the result accepted or rejected; a rejected report must not be treated as a clean result.
- The local Skill remains available for optional independent recomputation, offline runs, and report customization; it is no longer required for the cloud result. Research aid only; no target prices, position sizes, or buy/sell instructions.
Each card is one task. Fill in its fields below.
| As-of date (optional) | Ticker data packet (JSON string)* | Model assumptions (JSON string)* | Override model (optional; leave blank for default) | Actions | |
|---|---|---|---|---|---|
01 | |||||
02 | |||||
03 |