- News API subscription
- Scraping, parsing, and site-change maintenance
- Email verification service
- Waterfall email enrichment across 12 providers, plus gap-handling when providers miss
- Byline-resolution and outlet disambiguation logic
- Persistence for lists, jobs, review, sending, and tracking
Building your PR agent with ChatGPT? Right call. Now give it real data.
Keep ChatGPT for reasoning and drafting. Plug in Medialyst as the live news and verified-contacts layer through MCP where available, then use Newsjack.sh in a local agent when you need saved monitoring and the full PR workflow.
You are not choosing between ChatGPT and Medialyst. You are choosing what your ChatGPT agent can verify.
Vanilla ChatGPT vs. ChatGPT + Medialyst.
ChatGPT appears in both columns because the model is not the thing being replaced. The difference is whether it has live PR data, contact enrichment, and a durable workflow underneath it.
| Capability | Vanilla ChatGPT | ChatGPT + Medialyst |
|---|---|---|
| Strategy, critique, and pitch drafting | ✓ Strong at angles, positioning, critique, and first-draft pitches when you supply the evidence. | ✓ Same reasoning layer, grounded in live PR context before it drafts. |
| Agent workflow orchestration | ✓ Can coordinate prompts, files, skills, and tools that you wire into the workspace. | ✓ Gets PR-specific actions immediately: live news search, journalist enrichment, list persistence, and review. |
| Live news search with publication metadata | ~ ChatGPT can search pages, but metadata is sparse and inconsistent. | ✓ ChatGPT calls search_news for fresh results with source classification and publication context. |
| Maintained media database | ~ You own the expensive work: collecting publications, resolving journalists, refreshing beats, de-duplicating profiles, and cleaning stale contacts. | ✓ Medialyst gives your agent a maintained media database instead of making you build and refresh one in-house. Read why this is expensive to maintain |
| Journalist identity from live bylines | ~ Names may come from memory, snippets, or pages ChatGPT just found. | ✓ ChatGPT enriches from real article or author URLs and resolves the journalist behind the byline. |
| Email enrichment across 12 providers | ~ No maintained media database or 12-provider waterfall; guessed patterns and single-provider lookups leave gaps you chase manually. | ✓ Medialyst combines its media database with a 12-provider enrichment waterfall and real-time verification before an email reaches your list. |
| Pitch-fit scoring | ~ Good reasoning if you feed it current evidence. | ✓ Fit context and personalized angles grounded in observed work. |
| Date-windowed coverage checks | ~ Possible through manual prompting and search retries. | ✓ News queries can specify the publication window the agent needs. |
| Media-list persistence and re-enrichment | ~ Lives in chat files unless you build storage. | ✓ Lists, enrichment jobs, credits, review, send, and track live in the product substrate. |
| Hallucination risk on contacts | ~ Silent risk: a fake or stale contact still looks clean. | ✓ Lower risk because contacts are resolved from live inputs and verified before use. |
| Cost to first working agent | ~ Prompt prototype quickly; production data layer is yours to build. | ✓ Remote MCP plus Newsjack.sh gives the agent PR primitives immediately. |
| Maintenance burden | ~ You maintain scraping, parsing, enrichment providers, verification, and site drift. | ✓ Medialyst maintains the live news, enrichment, and verification substrate. |
Building the substrate is the expensive part.
The agent harness is not the cost center. The durable work is current news, journalist identity, deliverability, storage, and maintenance.
- News search: 0.1 credit per real-time query
- Journalist enrichment: 5 credits for complete contact info through a 12 provider waterfall, with real-time email verification
- Remote MCP: https://medialyst.ai/api/mcp
- Contract posture: No contract · no per-seat fees · 30-day money-back
Three layers. ChatGPT stays on top.
We open-sourced Newsjack.sh because the workflow should be inspectable. It handles PR motion, and it still needs live data underneath.
ChatGPT
Reasoning, drafting, critique, and orchestrationYour ChatGPT workspace, your prompts, your judgment. This is the interface layer for the PR agent.
Newsjack.sh
Open-source PR skills and local workflow runnerInstall with curl -fsSL newsjack.sh | bash. It adds the newsjack command plus skills for strategy, angle generation, journalist fit-checks, fact-checking, coverage tracking, saved monitors, and press clipping.
Medialyst
Live news and verified journalist data substrateHosted remote MCP at https://medialyst.ai/api/mcp, plus REST endpoints for real-time news search, journalist enrichment, and media-list infrastructure.
Connect ChatGPT to live PR data.
Once connected, ChatGPT can ground PR strategy, media lists, and pitch drafts in current news and verified journalist data instead of working from memory.
ChatGPT Apps or Developer Mode
Name:
Medialyst
Remote MCP server URL:
https://medialyst.ai/api/mcp
Authentication:
OAuth
Then enable Medialyst in the conversation tools menu.In the words of Medialyst users.
Real customers, approved for use. Your results depend on your story and your list — these are theirs.
"What came back was a working list with real journalists, current coverage, and contact information I could use the same day. No cross-referencing. No digging through three different tabs. One prompt and I was pitching."

"Medialyst makes it super easy to narrow down and create dedicated PR lists to reach out to the journalists that actually want to hear what you have to say. The timeline from chatting to the agent about what you have to promote to actually launching the campaign goes down to mere 1-2 hours instead of days worth of research and prospecting."

"Then recently I stumbled upon Elvis Sun's agentic PR tool Medialyst — with a few detailed prompts, the agent comes up with a pretty targeted list of journalists. The best thing is it gives a summary of their recent articles before I commit to reading every piece."

"Medialyst is the only tool that keeps up with the pace our firm needs to move at across our High Growth B2B Tech clients. The journalist targeting is sharp as well as accurate, and Elvis actually ships when you give him feedback."

"It's 100% customizable to you & your clients. It's affordable. It's built by someone who gets back to you quickly & truly wants to give the people what they want — MORE time."

"Medialyst takes the guesswork out of journalist outreach. It vets your pitch for suitability, matches it to the right journalists, and hands you their most accurate contact details — so instead of firing off cold emails into the void, you're landing in the right inboxes. The result? Fewer emails sent, more stories published."

"This is a must-have if you work in media, it's easy to use and saves you a lot of time! The customer service has been some of the best I have experienced."

Common questions about ChatGPT and Medialyst.
Can't ChatGPT just build a media list by itself?
Does Medialyst replace ChatGPT?
Can ChatGPT connect to Medialyst?
Do I have to build the agent workflow myself?
Why not just have ChatGPT search the web?
What does it cost to try?
Do you train on my queries or lists?
Is Medialyst affiliated with OpenAI?
Give your ChatGPT agent PR data you can trust.
Start with 300 free credits, connect Medialyst where your ChatGPT workspace supports MCP, and use Newsjack.sh in a local agent when you need saved monitoring or the full PR toolkit.