Perigon
Perigon provides an HTTP REST API for retrieving news and web content data, offering structured data suitable for various applications.
Verdict
Common use cases
- Monitor competitor mentions in trade press
- Track breaking news for crisis comms
- Gather sources for weekly industry briefing
- Pull sentiment data on brand coverage
- Research regulatory changes by jurisdiction
Integration
- Vendor
- Perigon
- Category
- other
- Auth
- API_KEY
- Composio slug
perigon
Tools
Setup
Setup guide
- 11. Sign up at Perigon's website and generate an API key from your account dashboard. 2. In Switchy, open your workspace settings and navigate to the Integrations tab. 3. Click 'Add Integration', search for Perigon, and select it. 4. Paste your API key into the authentication field and click 'Connect'. 5. Switchy will validate the key and confirm the connection. 6. Open any Space and type '@Perigon search articles about renewable energy policy' to test — you should see a list of recent articles with headlines and links. 7. If the response is empty, check your API key's rate limits and confirm your query matches Perigon's supported filters (topic, date range, source domain).
What teammates see: by default, memories from Perigon are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Daily Media Scan
@Perigon search for articles published in the last day about artificial intelligence regulation, limited to sources like Reuters, Bloomberg, and The Wall Street JournalOpen in a Space →
Competitor Intelligence
@Perigon find articles mentioning [Competitor Name] from the past week, and summarize the main themes and sentimentOpen in a Space →
Event Backgrounder
@Perigon search for articles about [Event or Topic] published in the last month, and list the five most relevant with summariesOpen in a Space →
Trend Spotting
@Perigon search for articles containing the phrase 'generative AI' published this week, grouped by source, and highlight any new anglesOpen in a Space →
Crisis Monitoring
@Perigon find all articles mentioning [Brand or Incident] published in the last six hours, sorted by recency, and flag any negative sentimentOpen in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@Perigon find recent news articles about renewable energy policy changes in the EU from the last 7 days
The MCP would return a structured list of articles matching the query — typically including headline, source publication, publish date, author, and a brief excerpt. For example: three articles from Reuters, Bloomberg, and The Guardian covering Germany's solar subsidy expansion and France's offshore wind auction results. Each entry includes a direct URL and relevance score based on the search terms.
This example shows the MCP's core news-search capability. Perigon aggregates content from thousands of sources, so results depend on API tier limits (free tiers may cap article count or freshness). The AI can synthesize findings across articles, but the MCP itself returns raw article metadata — no pre-digested summaries.
@Perigon search for mentions of our competitor 'Acme Corp' in tech industry coverage this month, then summarize sentiment trends
The MCP retrieves articles mentioning 'Acme Corp' from tech-focused publications over the past 30 days — perhaps 15 articles from TechCrunch, Ars Technica, and Wired. The AI then analyzes sentiment across headlines and excerpts: 60% neutral product coverage, 30% positive funding announcements, 10% critical pieces on customer support issues. The synthesis highlights a recent pivot to enterprise sales as the dominant narrative.
This demonstrates pairing Perigon's search with AI reasoning. The MCP provides article data; the AI interprets tone and patterns. Sentiment accuracy depends on excerpt length returned by the API — full-text access may require a paid plan. Rate limits apply to high-volume competitive monitoring.
@Perigon pull the top 5 articles about AI regulation from government and legal sources published yesterday, filter to US jurisdiction
The MCP returns five articles from sources like Federal Register announcements, Law360, and Bloomberg Government — each tagged with publication metadata and jurisdiction markers. Results might include an FTC workshop notice, a Senate subcommittee hearing summary, and three legal analysis pieces on state-level AI transparency bills. The AI can then extract key policy proposals or compare state approaches.
This example highlights Perigon's source-type and geographic filtering. The 'government and legal sources' filter narrows results to authoritative outlets, reducing noise. Jurisdiction tagging accuracy varies by article metadata quality. Users monitoring regulatory changes should verify official sources directly for compliance work.
Use-case deep-dives
When Perigon fits a 5-person leadership team's morning routine
A startup's exec team meets Monday and Thursday mornings to review industry news that might shift their roadmap. Before Perigon, someone manually scraped TechCrunch and The Information into a Notion doc. With Perigon's API, they pull curated articles filtered by topic and source into a shared Switchy thread each morning. The MCP handles authentication once, so the team doesn't pass API keys around Slack. This works well if your news needs are stable (same 8-10 sources, same 3-4 topics). If you need real-time alerts or custom RSS feeds that change weekly, Perigon's API structure gets rigid fast. For a predictable morning brief that saves 20 minutes of manual curation, this MCP is the right call.
How Perigon supports quarterly competitor research sprints
A 12-person product org runs a competitor-research sprint every quarter. They track 6 direct competitors and 4 adjacent platforms. Perigon's news aggregation pulls press releases, funding announcements, and feature launches into a single feed. The team uses Switchy to tag articles by competitor and product area, then summarizes findings in a Miro board. This setup works because the research is batched (not daily) and the competitor list is stable. If your competitive set shifts monthly or you need deep technical analysis beyond headlines, Perigon won't surface enough signal. But for a quarterly ritual where you need breadth over depth, the MCP keeps the team from drowning in Google Alerts and RSS chaos.
When Perigon helps a 3-person agency track client industries
A boutique agency manages 8 clients across fintech, healthcare, and logistics. Each account manager needs to stay current on their clients' industries without spending an hour a day reading. Perigon filters news by vertical and pipes it into client-specific Switchy channels. The team reviews articles during weekly check-ins and drops relevant links into client decks. This works if your client industries are broad enough that general news aggregation hits the mark. If a client operates in a niche subsector (like cold-chain pharma logistics), Perigon's sources won't go deep enough and you'll still need trade publications. For agencies with 5-15 clients in mainstream verticals, this MCP turns industry awareness from a chore into a 10-minute weekly habit.
Frequently asked
What does the Perigon MCP do in Switchy?
The Perigon MCP connects your workspace to Perigon's news and content API, letting your team query articles, headlines, and media sources directly from Switchy's chat interface. You can search by topic, source, date range, or keyword without leaving your workflow. It's useful for research, competitive intelligence, or content curation tasks where you need structured news data on demand.
Do I need a Perigon account to use this MCP?
Yes. You need an active Perigon account and an API key. Perigon issues keys through their dashboard after you sign up for a plan. Paste that key into Switchy's connection form and you're live. The key authenticates every request, so keep it private and rotate it if anyone leaves your team.
Can the Perigon MCP write or publish content?
No. Perigon is a read-only news API, so the MCP only fetches and filters articles. It can't post stories, update sources, or modify anything in Perigon's index. If you need to publish content based on what you find, you'll do that manually or through a separate integration like a CMS or social tool.
How is this different from just using Perigon's API directly?
The MCP wraps Perigon's API so your team can query news in plain English from Switchy's chat, without writing code or managing HTTP clients. You skip the boilerplate of parsing JSON responses and handling pagination. If you already have engineers building custom scripts, the API gives you more control; the MCP is faster for non-technical users.
Does using the Perigon MCP count against my Switchy plan limits?
The MCP connection itself doesn't consume extra seats or storage in Switchy. However, every query you run hits Perigon's API and counts against your Perigon plan's rate limits and monthly request quota. Check your Perigon dashboard to monitor usage and avoid overage charges if your team runs frequent searches.