Saddler examples

Stock Analysis Playbook

Turns a ticker list and investment question into an assumptions memo, sensitivity table, and source-linked report package.

Start from this example See more examples

Run teardown

One run produced these files, costs, limits, and follow-up choices.

InputTicker list with target horizon
Actions4 recorded highlights
Outputs3 produced files
LimitInvestment decisions still need analyst review
Source
Ticker list, valuation rubric, and market data
Output
Report, sensitivity table, source notes
Approval
No write approval needed; source freshness shown
Cost
$4.18 in 12m 06s

Input materials

  • Ticker list with target horizon
  • Investment question and audience
  • Approved valuation rubric
  • Previous outputs selected as context

Action log highlights

  1. Loaded the reusable valuation rubric and prior report context.
  2. Checked market-data availability before drafting assumptions.
  3. Built a sensitivity table and separated facts, assumptions, and open questions.
  4. Packaged the report with the source run for follow-up analysis.

Produced artifacts

Markdown

stock_report.md

Narrative report with assumptions, caveats, and next steps.

Produced by run_stock_demo_019 - Stock Analysis Playbook v1

CSV

sensitivity_table.csv

Scenario grid for valuation inputs and downside cases.

Produced by run_stock_demo_019 - Stock Analysis Playbook v1

JSON

source_notes.json

Source references and freshness notes used by the run result.

Produced by run_stock_demo_019 - Stock Analysis Playbook v1

Why this run matters

Research notes, model assumptions, and market data often split across spreadsheets and chat. The final report is hard to rerun because the source run and prompt are not preserved together.

How it was guided

The run used a short operating rule set. Keep the proof first, then open the prompt excerpt when you need to inspect the instructions.

Show prompt excerpt 
Analyze each ticker against the supplied investment question. Separate facts from assumptions. Cite source material, call out stale data, and produce a reusable report package with a sensitivity table.

What the agent could use

Connection

Market data access

Reads pricing and fundamentals through configured finance data connections.

Skill

Valuation rubric

Keeps comparable analysis, downside cases, and confidence labels consistent across runs.

Context

Investment policy

Applies the workspace's approved risk language and excluded recommendation types.

Memory

Prior coverage notes

Recalls only approved company notes and prior report caveats.

Cost, limits, next step

Run cost
$4.18
Duration
12m 06s
Human equivalent
About 90 minutes of analyst prep
Known limit
Investment decisions still need analyst review
Next action
Open the report package, adjust assumptions, or launch a follow-up run.

What we'd change

We would tighten the model-access posture for high-stakes finance workflows and require approval before adding any report to the shared Library.

Try it yourself

Start from this example in your workspace, launch it on the sample CSV, then open the run result before changing the prompt.

Start from this example