| name | weekly-deal-review |
| description | Parse meeting notes to generate a weekly deal pipeline review with prospect data, budgets, objections, and next steps. Works with Gemini Meet notes, Claap transcripts, or raw text. |
| user_invocable | true |
| arguments | [meeting_notes_path] [--format markdown|notion] [--range 7d|30d|all] |
Weekly Deal Review
Parse all meeting notes and produce a structured pipeline review — deal summary table + individual deal cards.
Inputs
| Argument | Required | Default | Description |
|---|
meeting_notes_path | No | $MEETING_NOTES_PATH env var, then ./meeting-notes/ | Directory containing .md meeting notes |
--format | No | markdown | Output format: markdown or notion |
--range | No | 7d | Time range: 7d, 30d, all |
Execution
Step 1: Locate Meeting Notes
1. Resolve notes path:
- Use argument if provided
- Else use $MEETING_NOTES_PATH env var
- Else use ./meeting-notes/
2. Glob for *.md files in that directory
3. If --range is set, filter by date:
- Extract date from filename (YYYY_MM_DD pattern) or first line of file
- Keep only files within the range
4. If no files found, report error and stop
Step 2: Parse Each Meeting Note
For each .md file, apply the parsing algorithm from references/meeting-parsing-guide.md:
Extract:
- date: from filename or first line
- prospect_company: from email domains + title + summary
- contacts: names and emails from Invités line
- source: how the lead came in (inbound/outbound/referral — infer from context)
- budget: amounts + context (monthly/one-time/per-unit)
- objections: list of concerns raised by prospect
- must_haves: features/requirements the prospect explicitly needs
- competitors: any competing solutions mentioned
- verbatims: direct quotes from the prospect
- next_steps: action items with owners and deadlines
- deal_stage: Discovery | Qualification | Proposal | Negotiation | Closed Won | Closed Lost
Step 3: Generate Pipeline Summary Table
## Pipeline Review — Week of [DATE]
| Prospect | Contact | Stage | Budget | Key Objection | Next Step | Risk |
|----------|---------|-------|--------|----------------|-----------|------|
| Company A | Name (email) | Qualification | €5-10K/mo | Platform recognition | Create account by Sat | 🟡 |
| Company B | Name (email) | Proposal | €2K | Content quality | Send brief | 🟢 |
Risk levels:
- 🟢 Low — clear next steps, engaged prospect, no blockers
- 🟡 Medium — some objections, timeline unclear, or missing decision maker
- 🔴 High — strong objections, competitor preference, stalled, or no budget
Step 4: Generate Individual Deal Cards
For each prospect, generate a detailed card:
### [Prospect Company] — [Stage]
**Contact:** Name <email> | **Date:** YYYY-MM-DD | **Source:** inbound/outbound
#### Budget
- [Amount and context]
#### Key Needs
- [Must-have 1]
- [Must-have 2]
#### Objections
- [Objection 1]
- [Objection 2]
#### Competitors
- [Competitor mentions with context]
#### Prospect Quotes
> "[Verbatim 1]" — Speaker Name
> "[Verbatim 2]" — Speaker Name
#### Next Steps
- [ ] [Action] — Owner — Deadline
- [ ] [Action] — Owner — Deadline
#### Deal Notes
[Any additional context, risks, or strategy notes]
Step 5: Output
Markdown (default):
- Print the full report to console
- If a file path is obvious (e.g., user's working directory), offer to save as
deal-review-YYYY-MM-DD.md
Notion (--format notion):
- Check if Notion MCP tools are available (
mcp__plugin_Notion_notion__notion-search)
- If available:
- Search for a "Deal Reviews" database or page
- Create a new page with the review content
- Return the Notion URL
- If NOT available:
- Fall back to markdown output
- Print: "Notion MCP not available. Output as markdown. To enable Notion, install the Notion MCP plugin."
Aggregation Insights
After all deal cards, add a summary section:
## Insights
- **Total active deals:** X
- **Total pipeline value:** €X - €Y (range based on budget mentions)
- **Deals needing follow-up this week:** X
- **Most common objection:** [theme]
- **Competitor landscape:** [most mentioned competitors]
- **Stalled deals (no activity > 14 days):** [list]
Error Handling
- No notes directory found → clear error message with path tried
- No .md files in directory → suggest checking path or date range
- Notes with no extractable data → skip with warning, don't fail the whole review
- Malformed dates → best-effort parsing, warn user