This guide explains the cleanest internal setup for leadership or operators who want quick reporting from the Deal Flow Matcher workflow, including questions like how many deals we scraped from Florida this week.
The recommended setup is simple: approved teammates get access to the private Deal Flow Matcher project, then use Codex from that workspace. This keeps reporting consistent with the production workflow and avoids shadow spreadsheets or duplicate logic.
| # | Name | Description |
|---|---|---|
| 1 | Private project access | Teammates should be added to the internal Deal Flow Matcher repository or equivalent internal workspace access path. |
| 2 | Codex access | They should open the project in Codex so Codex can inspect the scripts, run analysis, and answer reporting questions from the live workflow. |
| 3 | Approved local credentials | They need the organization-approved local access pattern for the workflow’s APIs. This should be handled through internal onboarding, not through shared docs or public pages. |
Once installed, teammates do not need to know the implementation details. They can ask for reporting directly in natural language.
How many deals did we scrape from Florida this week?
Show the count and list the matching deal titles.Show the top 10 states by scraped deal count for the last 7 days.Using title-based industry classification, how many HVAC,
Plumbing, and Landscaping deals did we scrape this month?Compare Florida deal counts for 2025 versus 2026
using Airtable Created At dates.Codex reads the existing workflow scripts, pulls the Airtable records, applies title-based enrichment when needed, and returns the requested report.
Faster answers, consistent definitions, and lower risk of people calculating the same metric in different ways.
| # | Name | Description |
|---|---|---|
| 1 | Pull data | Fetch Airtable deal records from the workflow source used by Deal Flow Matcher. |
| 2 | Filter | Limit by time period such as this week, this month, 2025, or 2026. |
| 3 | Group | Break out by state or by title-based industry classification. |
| 4 | Report | Return the answer in chat, spreadsheet form, or a leadership-ready summary. |
| # | Name | Description |
|---|---|---|
| 1 | Single source of truth | The same workflow logic powers both automation and reporting. |
| 2 | Less manual effort | Leadership can request reports without someone rebuilding filters by hand each time. |
| 3 | Consistent definitions | Questions like “Florida this week” are answered from the underlying data model, not a one-off spreadsheet interpretation. |
| 4 | Expandable | The same setup can support reports by state, time window, AE, title-based industry, or year-over-year trend. |