newsletter-signal-scanner

# Newsletter Signal Scanner

Turn your newsletter subscriptions into a structured intelligence feed. Monitors an AgentMail inbox for incoming newsletters, extracts signal-relevant content by keyword campaign, and delivers a weekly digest of what matters — competitor mentions, ICP pain language, market shifts, and emerging topics.

## When to Use

- "Monitor industry newsletters for competitor mentions"
- "Alert me when newsletters mention [topic] or [company]"
- "What are newsletters writing about this week in our space?"
- "Set up newsletter monitoring for [client]"

## Phase 0: Intake

### Newsletters to Monitor
1. Which newsletters should be subscribed to and monitored? (List names or URLs)
   - If unknown, ask: "What 3-5 newsletters does your ICP read?" — then use `sponsored-newsletter-finder` to discover others.
2. Which AgentMail inbox should receive them? (Or should we create a new one?)

### Keyword Campaigns
3. Competitor names to track (e.g., "Clay", "Apollo", "Outreach")
4. ICP pain-language terms to track (e.g., "outbound struggling", "pipeline dried up", "SDR ramp")
5. Market shift terms (e.g., "AI SDR", "agent-led growth", "GTM engineer")
6. Your brand name (to catch mentions)

### Output
7. Digest delivery: Slack channel, email, or markdown file? (default: markdown file)
8. Frequency: daily or weekly? (default: weekly)

Save campaign config to `clients/<client-name>/configs/newsletter-signals.json`.

```json
{
  "inbox_id": "<agentmail_inbox_id>",
  "keyword_campaigns": {
    "competitors": ["Clay", "Apollo", "Outreach", "Salesloft"],
    "pain_language": ["pipeline is down", "outbound isn't working", "SDR ramp"],
    "market_shifts": ["AI SDR", "GTM engineer", "agent-led"],
    "brand_mentions": ["YourCompany", "yourcompany.com"]
  },
  "newsletters": [
    {"name": "Exit Five", "from_domain": "exitfive.com"},
    {"name": "The GTM Newsletter", "from_domain": "gtmnewsletter.com"}
  ],
  "output": {
    "format": "markdown",
    "path": "clients/<client-name>/intelligence/newsletter-signals-[DATE].md"
  }
}
```

## Phase 1: Scan Inbox

Use the `agentmail` capability to fetch new emails from the monitored inbox:

```
Fetch emails from inbox <inbox_id> since <last_scan_date>
Filter to: known newsletter senders (match against newsletters config)
```

For each email:
- Extract subject, sender, date, full body text
- Strip HTML → plain text for analysis

## Phase 2: Apply Keyword Campaigns

For each newsletter email, scan for keyword matches:

```python
for email in emails:
    matches = {}
    for campaign, keywords in keyword_campaigns.items():
        found = []
        for keyword in keywords:
            if keyword.lower() in email.body.lower():
                # Extract context: 50 chars before + keyword + 50 chars after
                context = extract_context(email.body, keyword)
                found.append({"keyword": keyword, "context": context})
        if found:
            matches[campaign] = found
    email.signal_matches = matches
```

Only include emails with at least one keyword match in the digest.

## Phase 3: Extract Signal Snippets

For each matched email, extract clean signal snippets:

**Competitor mention example:**
> Newsletter: The GTM Newsletter | Date: 2026-03-05
> Campaign: competitors
> Keyword: "Clay"
> Context: "...teams that use **Clay** for enrichment are seeing 3x better personalization rates compared to..."

**Pain language example:**
> Newsletter: Exit Five | Date: 2026-03-04
> Campaign: pain_language
> Keyword: "outbound isn't working"
> Context: "...a lot of founders telling me **outbound isn't working** the way it used to. The reply rates I'm seeing..."

## Phase 4: Output Format

```markdown
# Newsletter Signal Digest — Week of [DATE]

## Summary
- Newsletters scanned: [N]
- Emails with signals: [N]
- Top trending topic: [topic]

---

## Competitor Mentions

### Clay
- **[Newsletter Name]** — [Date]
  > "[Context snippet]"
  Source: [email subject] | [URL if available]

### [Other Competitor]
...

---

## ICP Pain Language

Signals suggesting your ICP is feeling pain your product solves:

- **[Newsletter Name]** — [Date]
  > "[Context snippet]"
  — Relevance: [why this matters]

---

## Market Shift Signals

Emerging topics gaining newsletter coverage:

- **"[Topic]"** — mentioned in [N] newsletters this week
  > "[Context snippet]"

---

## Your Brand Mentions
[Any mentions of your company or product]

---

## Recommended Actions
1. [Specific action based on signals — e.g., "Exit Five is covering AI SDR fatigue — good moment to publish our take"]
2. [Competitive response if needed]
```

Save to `clients/<client-name>/intelligence/newsletter-signals-[YYYY-MM-DD].md`.

## Phase 5: Setup — Subscribe to Newsletters

For first-time setup, subscribe the AgentMail address to target newsletters:

1. Get the AgentMail inbox address (via `agentmail` capability)
2. For each newsletter, visit subscription page and submit the AgentMail address
3. Confirm subscriptions (check inbox for confirmation emails)
4. Allow 1-2 weeks of accumulation before first full digest

## Scheduling

Run weekly (Monday morning recommended):

```bash
# Every Monday at 7am — before the team's standup
0 7 * * 1 python3 run_skill.py newsletter-signal-scanner --client <client-name>
```

## Cost

| Component | Cost |
|-----------|------|
| AgentMail inbox | Depends on AgentMail pricing |
| Email parsing + keyword matching | Free (local logic) |
| **Total** | **Near-zero ongoing cost** |

## Tools Required

- **AgentMail API** — for inbox access
- **Upstream skill:** `agentmail` capability

## Trigger Phrases

- "Scan newsletters for this week's signals"
- "What are industry newsletters saying about [topic]?"
- "Run newsletter signal scanner for [client]"
- "Set up newsletter monitoring"