Category: Health

It’s getting real out here.

Ever since I shared how autonomous AI Agents can monitor and heal the network on their own—yes, that one—I’ve gotten the same follow-up question in different forms:

“Okay, Kareem, this all sounds great… but how do I actually build one of these Model Context Protocol (MCP) servers for my product?”

Good news! If your product—like almost every product out there—has APIs, then chances are that you already have what you need.

Enter: OpenAPI spec

OpenAPI is a friendly contract for your APIs. You might’ve used it for Swagger docs, SDKs, Postman collections, or that one dusty codegen project from 2021. But here’s the twist: What if you passed that same OpenAPI spec to your AI agent?

That’s it. That’s the key.

One OpenAPI spec → one MCP Server → one AI-powered, access-controlled gateway to your product.

And no, this isn’t a “12 steps and a DevRel miracle” situation. It’s just a few lines of Python and a FastMCP wrapper around your OpenAPI file. The magic? Your APIs get transformed into safe, role-based AI tools—without writing a single custom tool definition.

Consider the following example:

You’re wrapping your existing OpenAPI spec with FastMCP, wiring in your authenticated client, and passing in your route-based ACLs. That’s how simple it is to go from “API docs” to “AI-ready, access-controlled MCP server.” 

Build fast, govern smart

In this new AI-powered world, speed is the easy part. Governance—that’s the harder lift.

We don’t want to give the agent the keys to the kingdom. We want to provide it with a badge with just the proper access.

That’s where RouteMap comes in—our ACLs for AI. With a simple list of patterns (regex for those who love pain and suffering) and HTTP verbs, you can declare what endpoints are accessible for different personas (NOC, Sysadmin, full access, and so on).

Yes, it’s literally that easy. You’re building endpoint ACLs as code. You don’t need to create a whole new auth system or train a model to “learn” permissions. You just declare what roles get access to what endpoints—and the MCP Server enforces it.

From chaos to order

Let’s walk through a real-world use case.

Say you’re a NOC team managing a multi-site Meraki deployment. You’re responsible for keeping network devices patched and secure—but you can’t just schedule firmware upgrades at any time. Some sites are 24/7. Some spike at noon. Some run night shifts. The ideal upgrade window is a moving target.

That’s where the agent steps in.

You want to give the agent just enough access to help:

• Pull the current firmware status
• Monitor network usage patterns
• Schedule upgrades when it makes sense

Meanwhile, your Sysadmin team needs the agent to generate compliance reports. They need to know which devices are running outdated firmware—but they’re not scheduling upgrades or touching live traffic.

Two personas. Two very different scopes. One MCP server.

Here’s the beauty of it all. We didn’t write any custom tools. We didn’t build workflows or hardcode business logic. We just fed the MCP server the full Meraki OpenAPI spec—and let RouteMap handle the rest:

The NOC agent can schedule upgrades, because it needs that control. The Sysadmin agent? It gets a read-only view, tailored for visibility and compliance.

And again—we didn’t tell the agent how to do anything. The magic is in the MCP server. The tools become available based on the role, and the AI figures out the rest.

That’s the kind of governed autonomy that turns AI from a risk into a capability.

View it in action

As usual, you’ll find everything I’m showing here—the MCP server code, config, and prompt—in my GitHub Repo.

Now let’s fire this thing up. (And, yes, Network Pharaoh is a thing now.)

With the MCP server running and our route maps defined, I launch Claude Desktop (my MCP client of choice) and type the following prompt:

Your name is Network Pharaoh. You are acting with complete administrative visibility and data access privileges. You are a senior network administrator overseeing multiple Cisco Meraki organizations across the enterprise. Your role is to ensure that all network devices are running the latest compliant firmware. You are authorized to recommend firmware upgrades, but you must wait for explicit human approval before initiating any updates.

Target Organizations: Cisco U.

Task Instructions – For each organization:

• • List all networks
  • For each network, list all associated devices (including model, serial, and current firmware version)
  • Retrieve the available firmware upgrade recommendations for the organization
  • Identify any device that is not running the recommended version
  • Recommend firmware upgrades as appropriate
  • Do not perform any upgrade unless the human explicitly confirms with a statement like: “Yes, please upgrade [device/network].”

A few things are worth calling out:

The human-in-the-loop is built in. The agent knows it can’t act on its own—it must wait for approval. That’s governance baked into the prompt.

We didn’t tell the agent how to check compliance or suggest upgrades. It uses the tools available through the MCP Server and acts within the boundaries defined by its role.

The agent is doing intelligent work inside safe boundaries—using only what it’s been given access to. No guesswork. No scraping. No uncontrolled API calls. Just clean, policy-driven interaction through a structured, secure interface.

Here’s what the MCP server config looks like behind the scenes:

Pay attention to the importance of the MCP_ROLE. This one environment variable controls which routes the agent has access to. Set it to “NOC” and the agent can recommend firmware upgrades. Set it to “sysadmin” and the same agent, with the same prompt, will only be able to generate compliance reports—no upgrades, no PUTs.

That’s the advantage of separating the intelligence (LLM) from the control plane (MCP). You stay in charge of what the agent can do.

And here’s what the MCP server makes happen:

1. Network Pharaoh traverses our Cisco U. organization, pulling a list of managed devices and spitting out a report.
   
2. As Network Pharaoh is waiting for a human in the loop to execute the upgrade, it also auto-corrects the version based on web search and schedules it for us based on usage.
   
3. Et, voila!
   

The skills behind the scenes

Let’s zoom in for a second. What did it take to build this?

Here are the skills a network engineer needs to put this together:

• Understanding of API fundamentals: OpenAPI specs, endpoints, HTTP methods
• Python scripting: Spinning up a basic server and configuring the MCP wrapper
• Access control thinking: Defining roles, access boundaries, and enforcing least privilege
• Agent design mindset: Prompting with context, purpose, and clear human oversight
• Curiosity and experimentation: Trying things out and tweaking as you go

And maybe most importantly:

• A shift in thinking—from building automation for the network, to building automation that understands the network.

Let’s keep pushing this frontier. Because the more we build intelligent boundaries, the more we unlock safe autonomy.

And that’s how we go from the Wild West… to a well-governed AI-powered enterprise.

 

Sign up for Cisco U. | Join the  Cisco Learning Network today for free.

Learn with Cisco

X | Threads | Facebook | LinkedIn | Instagram | YouTube

Use  #CiscoU and #CiscoCert to join the conversation.

 

Read more from the AI Break series:

Share:

To save the lives of infants and small children living in low- and middle-income countries, there are a handful of tried and tested tools, like anti-malarial drugs, bed nets and vaccines. The results from a massive experiment in rural Kenya suggests another: cash.

Infants born to people who received $1,000, no-strings-attached, were nearly half as likely to die as infants born to people who got no cash, according to a report published Monday by the National Bureau of Economic Research. Cash cut mortality in children under 5 by about 45%, the study researchers found, on par with interventions like vaccines and anti-malarials.

“This paper is really well done, and the result itself is pretty stunning,” says Heath Henderson, an economist at Drake University who wasn’t involved in the study. Historically, it’s been “difficult to study the impacts of cash transfers on mortality with any sort of rigor,” he says. “This study is different,” he says, and suggests cash can help people get life-saving care.

Over the past decade or so, the idea of simply giving people living in poverty cash has gained traction, in part by evidence that it can work. The best evidence comes from what researchers call randomized controlled trials. In this set-up, an experimental group gets cash, a control group doesn’t, and researchers look for differences in measurable outcomes, like income or savings, to understand what difference cash made.

While studies have found clear links between cash transfers and economic well-being, health has been harder to pin down, especially for the most dire health outcomes.

“Infant and child mortality in rural Kenya is an order of magnitude higher than it is in the U.S.,” says Edward Miguel, an economist at the University of California Berkeley and study co-author. “But it’s still a relatively rare event to have a child die. Statistically speaking, that means we need a really large sample size to have precise and reliable estimates of the effect of cash on child mortality.”

$1,000 to 10,000 families

In 2014, the nonprofit GiveDirectly began a massive experiment. Over the next three years, they gave $1,000 to over 10,000 low-income households across 653 villages in Western Kenya.

“It was designed as a randomized control trial,” says Miguel. “So some areas got more cash. Some got less cash, and we can study the impact of that cash.”

To study that impact, Miguel and his colleagues collected a lot of data. They completed a kind of birth census for all children that had been born and died before age 5 over the previous decade in the study area. “We ended up collecting data on over 100,000 births. It took a year to do.”

They found that cash had major benefits for infant and child mortality, especially when it was delivered close to birth.

Cash payments were associated with a 48% drop in infant mortality, from roughly 40 deaths per 1,000 births to about 21 deaths. Deaths of children under five were 45% lower in households who got cash, dropping to 32 per 100,000 births from 57.

Cash played an outsized role in reducing deaths during birth and in the few weeks after, falling by 70% compared to controls. “That really pointed toward a key role for access to health services right at the moment of delivery being very important,” says Miguel.

Why cash cut deaths

For many living across rural sub-Saharan Africa, getting to a health facility, and paying for care there, can be difficult, especially when pregnant.

“When I worked in rural parts of Uganda, one of the things that was really clear for pregnant women was they did not attend antenatal care, because it’s so difficult to get to a health care facility,” says Miriam Laker-Oketta, GiveDirectly’s senior research adviser.

“You’re making the decision between, should I go for antenatal care and have my family sleep hungry, or should I stay home and hope that my baby is fine because I’m not feeling sick and we can have a meal that day,” she says, since often women would have to forgo work for a day to go to the doctor. “Those are some of the decisions people have to make.”

Extra cash seems to make those decisions easier, as long as health care facilities weren’t too far away.

Cash made the biggest difference for families who live roughly 30 minutes or less away from a health care facility staffed with physicians. When the distance is greater, the benefits of cash for infants start to wane, though do not disappear entirely.

The researchers saw 45% more hospital deliveries among pregnant people who received cash than those who didn’t. It’s often more expensive to deliver at a hospital than a smaller clinic, says Laker-Oketta. “We’ve given people the means to access the care that they need and not to make some of these really difficult choices between getting care and feeding a family.”

The extra cash also helped put more food on the table. Children were about 44% less likely to go to bed hungry in households that received cash, the study found. Women who got cash while pregnant also worked about half as much — roughly 21 fewer hours per week — in their first trimester and the months after delivery than women who didn’t get cash. Work in these rural areas can often be physically taxing, says Laker-Oketta.

“That’s great for the mother’s health, but also gives time for her baby to develop well,” she says. “She’s also available after the baby is born to take the child to any early health visits.”

A ‘very important’ data point

Altogether, the results impressed Aaron Richterman, a physician who studies poverty reduction at the University of Pennsylvania and wasn’t involved in the study.

“It’s one data point, but it’s a very important data point. We can be very certain that in this case, the cash caused these benefits in mortality that we’re seeing,” he says. In an environment of shrinking foreign aid, he says cash could offer a simple way of reducing infant mortality.

Just how big a difference cash could make may depend, in part, on how readily people in other locations can use the extra money to get health care.

“I think this paper underscores the point that it’s really adequate access to health care that’s making all the difference,” said Henderson, the Drake economist and author of the book Poor Relief: Why Giving People Money Is Not The Answer To Global Poverty. “It just so happens that in this particular place, people needed cash to access health care.”

That’s likely the case in many places across sub-Saharan Africa after years of investment in bolstering health care systems, says Laker-Oketta, but not all.

“The answer is not we give cash alone, or we just focus on improving the health care system,” she says. “What’s obvious in this study is that you need both to be working together.”

When the seasons change, or when I’ve been around a lot of the sniffles, I like to make a big batch of elderberry syrup. While that’s always been a staple around our home, I also like to change up my immune boosting remedies. I’ve made all sorts of herbal concoctions over the years, but this simple ginger, turmeric, and lemon immunity shot is one I keep coming back to.

It’s warming and a little spicy, with a vibrant golden color thanks to the antioxidants. Best of all, it’s packed with nutrients that support the immune system, calm inflammation, and help the liver with detox. I sometimes sweeten it with raw honey to take the earthy edge off of the spices. If you’re feeling fancy, swap out the water for coconut water or pineapple juice for a tropical twist.

Why I Love These Ingredients

Fresh Ginger

Ginger is a staple in my kitchen for both cooking and natural remedies. I use it for everything from easing upset stomachs to calming coughs. It’s naturally warming, supports healthy circulation, and contains compounds that can help reduce inflammation. If you eat ginger you can feel it working almost immediately to warm the body. 

Fresh Lemon Juice

Fresh lemon juice adds a bright flavor and a welcome dose of vitamin C. It’s great for supporting immune health and can help soothe a scratchy throat. I also find that lemon gives this drink a refreshing balance. It cuts through the spice of the ginger and the earthiness of the turmeric to brighten it up. Lemon is also a great way to detox. 

Fresh Turmeric

Many recognize turmeric for its anti-inflammatory and antioxidant properties. I love using it to support joint health, digestion, and liver function. Fresh turmeric root works great for this recipe and lends an earthy, complex flavor and a bright yellow hue. Just remember, it will stain anything it touches, so be mindful when you’re making the immunity shots. 

Raw Honey 

The raw honey is optional, but it does help to mellow out the bite from the ginger and turmeric. Honey has natural antibacterial properties, can soothe sore throats, and adds a mellow sweetness. Local, raw honey can help combat seasonal allergies and is high in antioxidants and nutrients. It can even help reduce mucus and calm coughs as well as some over the counter cough syrups.

Optional Base Swap

While plain water works just fine, sometimes I’ll use coconut water for extra electrolytes. Pineapple juice is another great option for natural sweetness. Pineapple contains the enzyme bromelain to help better digest food and calm coughs. Other research shows bromelain can help with issues like preventing sinusitis, bronchitis, and blood clots. Both options pair beautifully with the ginger and turmeric.

Here’s how to make your own anti-inflammatory, immune boosting wellness “shots.”

A Quick Tip on Turmeric

Turmeric is wonderful, but it’s also notorious for staining countertops, wooden spoons, and anything light-colored. I’ve learned the hard way to wipe up any splashes right away.

Final Thoughts On Immunity Shots

This ginger, turmeric, and lemon immunity shot is a great way to keep your immune system strong all year long. It’s easy to make and tastes delicious. I’ve found that it’s a little too strong for younger children. However, you could try diluting in more water (or pineapple juice) and mixing more honey to mellow the flavor for little ones.

To boost the benefits even more, try stirring in a pinch of black pepper before serving! This helps your body absorb more of the curcumin from the turmeric. Play with the ingredients some and make it your own!

What are some of your favorite, easy ways to get a health boost? Leave a comment and let us know!

Public safety, emergency vehicle prioritization, and reducing urban congestion are but a few of the goals of intelligent Transportation Systems (ITS). ITS provides real-time traffic incident detection, connected traffic lights, and integrated public transit data, potentially saving billions annually in lost productivity and fuel, and supporting key sustainability and compliance initiatives.

At Cisco, our industrial networking and security solutions are driving the modernization of roadways by enabling the next generation of Intelligent Transportation Systems (ITS). Our solutions provide the secure, reliable, and scalable connectivity essential for putting roadside data to work – improving safety for all, with real-time roadside responsiveness, and operational efficiency.

We will be at the ITS World Congress 2025 in Atlanta (Aug 24-28) to showcase how we can help you build a durable ITS networking backbone that provides resilient connectivity for digital signage, digital roadways, traffic management systems, and pedestrian safety. Be sure to check out our real-time interoperability demonstration across multiple vendors on the exhibition floor, featuring Q-Free’s Kinetic Mobility ATMS integrating live intersections, RSUs, and live camera feeds using a Cisco network.

Visit us at Booth #1741 we’d love to hear where you are on your roadway’s modernization journey.

Let’s talk briefly about how we can help:

Digital Signage: Enhancing Traffic Communication and Data Utilization

Modernized roadways depend on dynamic communication. Digital signage, connected by Cisco’s industrial networking solutions, delivers real-time messages about traffic conditions, hazards, and route options to drivers and pedestrians. By utilizing secure connectivity-even in remote or harsh environments-these systems ensure that critical messages are transmitted without interruption. Through collaboration with leaders like Daktronics, Cisco enables digital signs to both receive up-to-the-minute updates and transmit critical situational data back to traffic management centers, thus transforming roadside data into actionable insights while maintaining the highest standards of cybersecurity.

Digital Roadway: Securing and Connecting the Modern Infrastructure 

The digital roadway is at the heart of roadway modernization, interconnecting traffic signal controllers, cameras, weather sensors, and more. Cisco’s ruggedized, space-efficient devices withstand extreme roadside conditions (tested to National Electrical Manufacturers Association, [NEMA] TS2) while providing seamless, secure connectivity. With SD-WAN and its configuration catalog, you can deploy quickly and at scale. To combat the rise of cyber threats targeting transportation infrastructure, Cisco integrates advanced cybersecurity features-including Next-Generation Firewall (NGFW), intrusion detection/prevention (IDS/IPS), and network segmentation-directly into its industrial routers and switches.

In addition, Cisco industrial routers give you visibility into connected roadway assets and their security posture to help you secure your operations. They also come with zero trust remote access capabilities to empower remote experts and secure their activities. Encrypted tunnels (IPsec) further safeguard the vast amounts of data generated and shared by these connected devices, ensuring that modernization does not come at the expense of security. As a result, agencies can confidently leverage roadside data to optimize operations and respond to incidents in real time.

Traffic Management Systems: Real-Time Efficiency Through Data and Security

Cisco’s networking enables traffic management systems, including those from Q-Free, to leverage real-time, data-driven control of signals and roadside devices. Modern roadways require connected infrastructure that not only supports remote management and dynamic signal adjustments but also collects a wealth of operational data. This data, protected by Cisco’s robust security framework, feeds into AI and analytics platforms for optimizing traffic flow, prioritizing emergency vehicles, and reducing congestion.

Too often, allowing 3^rd party access to ITS systems results in a high degree of security exposure. With Cisco’s secure remote access to roadside cabinets for contractors and vendors, 3^rd party access can be securely allowed for a virtual truck roll – resulting in fewer physical maintenance visits and greater operational resilience, all while protecting critical systems from evolving cyber threats.

Pedestrian Safety: Data-Driven Protection for Vulnerable Road Users

Modern roadways prioritize the safety of all users, especially pedestrians. Cisco’s connected infrastructure integrates video analytics and smart detection (radar/lidar/microwave detection), ensuring that data from cameras and sensors is used to adapt signal timings and enhance intersection safety. Reliable, secure connectivity means these systems operate around the clock and are shielded from cyber threats, supporting real-time responses and long-term safety improvements. By putting roadside data to work, transportation agencies can identify risk patterns and implement targeted interventions to protect pedestrians and other vulnerable road users.

Conclusion: A Foundation for the Future of Roadways

To make transportation systems intelligent you need secure and resilient connectivity. Cisco’s industrial networking and security solutions are essential to the ongoing modernization of roadways-transforming legacy infrastructure into intelligent, data-powered, and cyber-resilient transportation networks. Our networking solutions for ITS bring the expertise and innovation of enterprise networking into our secure, industrial networking – all proven in ITS environments. Cisco enables transportation agencies to harness the full value of their roadside data. The result is safer roads, more efficient operations, and a robust foundation for future innovations such as connected and autonomous vehicles.

Subscribe to Cisco Industrial IoT Newsletter

Share:

From pricey wearable devices to your phone, it has never been easier to track your daily physical activity, or lack of it. And if you’re like many Americans — and spend 9-plus hours sitting everyday — chances are you could probably stand to take a few more steps.

But just how many should you aim for if you want to live a longer and healthier life? New research suggests 7,000 is a good target.

Scientists have made big strides in the science of step-taking over the past decade, gathering tons of data.

And no surprise, they find the more you move, the less likely you are to die from cardiovascular disease and other ailments.

But, they’ve also concluded that the widely-cited target of 10,000 steps is not rooted in solid science — it grew out of a Japanese promotional campaign.

In fact, a number of studies have now landed on a new number to keep in our heads as we trudge through the day.

“7,000 steps tends to be the range where there seems to be diminishing return on investment for increasing more steps,” says Melody Ding, a professor of public health at the University of Sydney.

Ding is the latest researcher to dive into the evidence, as part of an effort to update the physical activity guidelines in her country.

Her team reviewed data collected from more than 160,000 adults around the world on the link between step count and a variety of health outcomes.

Among their findings: Taking 7,000 steps per day was associated with nearly a 50% lower risk of dying compared to the bare minimum of 2,000 steps.

The study, published in the Lancet Public Health, also showed the chance of developing cancer fell by 25%, type 2 diabetes 14%, cardiovascular disease 25%, symptoms of depression 22% and dementia 38%.

Now, Ding says there were still some improvements for those who exceeded 7,000 steps, but overall they saw “diminishing returns on investment,” meaning the added health benefits across the population become smaller and smaller.

“It definitely doesn’t do harm if you go beyond 7,000,” she says, “So for anyone who’s already doing 10,000 and more, there is no point of going back, but for the folks who are finding it harder to get there, 7,000 could be a really realistic target.”

The research underscores that, on the lower end, even a modest jump in steps can have a major impact on your well-being.

For example, simply jumping from 2,000 to 4,000 steps in a day was accompanied by nearly a 36% lower risk of dying.

The work from Ding’s team reflects a growing consensus in the scientific literature that around 7,000 steps could be a reasonable target for the public. (There is some variation in the exact number depending on the study.)

Amanda Paluch, who has published two meta-analyses on step count and health, has found a similar range as this latest study, though her work suggests the step targets may also vary with age.

“We were seeing that older adults did not seem to need as many steps compared to younger adults,” says Paluch, a physical activity epidemiologist at the University of Massachusetts Amherst.

Specifically, Paluch’s research shows the decreased risk of dying appeared to plateau between 6,000 to 8,000 steps for people in their 60s and older, versus 8,000 to 10,000 for younger adults.

She says this latest study is one of the first to consider a broader array of health outcomes than just mortality and cardiovascular disease, but cautions some of the results, such as those related to cancer, depression and dementia, are still somewhat “preliminary” because they’re based on only a few studies.

Researchers have also tried to pin down whether speed makes a difference. Here, the new Lancet study could not make any definitive conclusions, in part because there are various ways to measure intensity and differences could simply reflect better overall fitness and physical function.

“We actually don’t see an association once we consider the total number of steps,” says Paluch, who has also looked at this question. “So, essentially, the total number of steps, regardless of how fast you’re walking, seems to have a benefit,” she says.

The federal government’s current physical activity guidelines don’t actually recommend a daily step target. Instead, they focus on time, recommending 150 to 300 minutes of moderate intensity exercise a week for adults, or 75 minutes to 150 minutes of vigorous exercise.

But Dr. William Kraus, who worked on those guidelines, would love to see steps included the next time they are updated.

“Because they’re objective measures that anyone can get off their phone, right?” says Kraus, a cardiologist whose lab studies exercise and health at Duke University.

He says ultimately both approaches to measuring physical activity reflect the same thing — the amount of energy being burned up through movement.

“What you’re seeing is the more energy expenditure that you consume with physical activity, the greater benefit you get,” he says.

And because the point is to get physical activity – and expend energy – it doesn’t have to be just steps. Kraus says you can translate one mile of walking into one fifth of a mile swimming, or five miles cycling.

Melody Ding at the University of Sydney says she likes to think of movement in the same way as diet.

“Just like we need a balanced, healthy diet, it’s important for us to have balanced physical activity, training throughout the week as well,” she says.

In other words, get those steps, but also make time for mobility and strength training. All of it adds up.

If you are counting steps, Kraus emphasizes that the evidence pointing to roughly 7,000 steps a day should not discount the value of doing more steps, even if the upsides that have been studied so far become incremental above that number.

“Everybody wants to know how little I need to do. That is the wrong question,” he says, “Anything is better than nothing, more is better than less.”

What’s more, the calculations appear to change when you factor in how much time people spend sitting.

He cites evidence that suggests people need to aim higher than 7,000 or even 10,000 if they spend eight hours in a chair.

“If you can get 13,000 steps, you can get rid of all the negative risks that are associated with sitting, ” he says.

Our family prioritizes healthy, whole foods, so why not do the same for our pets? To my children, pets truly are part of the family. It was their idea to make homemade dog treats for our dog. And now they wanted to make homemade cat treats for our cats, Penelope and Tiger, too (“to be fair”).

Like the dog treats we made for Daisy, these homemade treats contain healthy ingredients cats love. They’re simple to make and store really well in the fridge. Our cats don’t even seem to mind if we don’t heat them up before giving them one. 

These treats are part of the Natural Cat Care & Holistic Alternatives we use to keep our furry friends healthy. Just remember these are treats though and not intended to be a meal replacement. 

Homemade Cat Treats: What You Need

There are so many ways you could make this recipe your own. I’d love to hear how you adapt it for your pets in the comments. Here’s what we used and why:

Coconut Oil

Coconut oil is a definite staple at our house, for both humans and pets. For cats, it’s a great source of healthy fats, especially the medium-chain fatty acids that support digestion and healthy skin. I use it in these homemade treats since coconut oil has natural antibacterial and antifungal properties. 

Cats are naturally carnivores so their systems are designed for fat-based energy. Coconut oil is a nourishing addition that mimics the natural fats they’d get from prey in the wild.

Coconut Flour

Coconut flour is one of my favorite grain-free alternatives when I’m baking for the family. So it made sense to use a little in treats for our cats too. It’s high in fiber, gentle on digestion, and helps everything hold together without using grains. Plus, it blends really well with the coconut oil and eggs, creating a soft texture our cats love.

Cooked Sweet Potatoes

While cats don’t need carbs, sweet potato adds some extra nutrients, especially vitamin A and fiber. It also gives a soft, easy-to-digest base for the treats. I’ve found that when used in moderation, it can be a nice way to add variety and a soft texture. 

Gelatin Powder

 I use gelatin in a lot of homemade recipes, from marshmallows to gummies. It turns out it’s great for pets too. Gelatin is rich in amino acids like glycine and proline, which support healthy skin, fur, and joints. I like to use a bulk grass-fed gelatin to keep it budget-friendly, but any clean gelatin will work. It also helps bind the treats and gives them a chewy texture our cats enjoy.

Tuna (or Sardines)

When in doubt, tuna is always a hit! Most cats can’t resist it. Canned fish like tuna or sardines are not only protein-rich, but they’re full of omega-3 fatty acids that support brain, skin, and coat health. In our house, sardines are a favorite. They’re small, have less mercury, and are packed with nutrients.

Egg

Eggs are a complete protein source, and full of healthy fats and fat-soluble vitamins. I use them in homemade pet treat recipes because they help everything hold together while adding a big nutritional boost. If your cat tolerates eggs well, they’re a simple and affordable way to add some extra nourishment.

I baked these on a baking sheet with natural parchment paper to avoid sticking. They store well in a glass jar in the fridge.

Quality Cat Food

Many of you have asked what cat food we use. I’ve tried different brands over the years and even experimented with making our own. Ultimately, I prefer to leave this in the hands of the experts. We now alternate between a variety of grain-free cat foods (wet, never dry). 

Have you ever made homemade treats for your pets? What are their favorite ingredients? Leave a comment and let us know!

Alright, my friends, I’m back with another post based on my learnings and exploration of AI and how it’ll fit into our work as network engineers. In today’s post, I want to share the first (of what will likely be many) “nerd knobs” that I think we all should be aware of and how they will impact our use of AI and AI tools. I can already sense the excitement in the room. After all, there’s not much a network engineer likes more than tweaking a nerd knob in the network to fine-tune performance. And that’s exactly what we’ll be doing here. Fine-tuning our AI tools to help us be more effective.

First up, the requisite disclaimer or two.

1. There are SO MANY nerd knobs in AI. (Shocker, I know.) So, if you all like this kind of blog post, I’d be happy to return in other posts where we look at other “knobs” and settings in AI and how they work. Well, I’d be happy to return once I understand them, at least. 🙂
2. Changing any of the settings on your AI tools can have dramatic effects on results. This includes increasing the resource consumption of the AI model, as well as increasing hallucinations and decreasing the accuracy of the information that comes back from your prompts. Consider yourselves warned. As with all things AI, go forth and explore and experiment. But do so in a safe, lab environment.

For today’s experiment, I’m once again using LMStudio running locally on my laptop rather than a public or cloud-hosted AI model. For more details on why I like LMStudio, check out my last blog, Creating a NetAI Playground for Agentic AI Experimentation.

Enough of the setup, let’s get into it!

The impact of working memory size, a.k.a. “context”

Let me set a scene for you.

You’re in the middle of troubleshooting a network issue. Someone reported, or noticed, instability at a point in your network, and you’ve been assigned the joyful task of getting to the bottom of it. You captured some logs and relevant debug information, and the time has come to go through it all to figure out what it means. But you’ve also been using AI tools to be more productive, 10x your work, impress your boss, you know all the things that are going on right now.

So, you decide to see if AI can help you work through the data faster and get to the root of the issue.

You fire up your local AI assistant. (Yes, local—because who knows what’s in the debug messages? Best to keep it all safe on your laptop.)

You tell it what you’re up to, and paste in the log messages.

After getting 120 or so lines of logs into the chat, you hit enter, kick up your feet, reach for your Arnold Palmer for a refreshing drink, and wait for the AI magic to happen. But before you can take a sip of that iced tea and lemonade goodness, you see this has immediately popped up on the screen:

Oh my.

“The AI has nothing to say.”!?! How could that be?

Did you find a question so difficult that AI can’t handle it?

No, that’s not the problem. Check out the helpful error message that LMStudio has kicked back:

“Trying to keep the first 4994 tokens when context the overflows. However, the model is loaded with context length of only 4096 tokens, which is not enough. Try to load the model with a larger context length, or provide shorter input.”

And we’ve gotten to the root of this perfectly scripted storyline and demonstration. Every AI tool out there has a limit to how much “working memory” it has. The technical term for this working memory is “context length.” If you try to send more data to an AI tool than can fit into the context length, you’ll hit this error, or something like it.

The error message indicates that the model was “loaded with context length of only 4096 tokens.” What is a “token,” you wonder? Answering that could be a topic of an entirely different blog post, but for now, just know that “tokens” are the unit of size for the context length. And the first thing that is done when you send a prompt to an AI tool is that the prompt is converted into “tokens”.

So what do we do? Well, the message gives us two possible options: we can increase the context length of the model, or we can provide shorter input. Sometimes it isn’t a big deal to provide shorter input. But other times, like when we are dealing with large log files, that option isn’t practical—all of the data is important.

Time to turn the knob!

It is that first option, to load the model with a larger context length, that is our nerd knob. Let’s turn it.

From within LMStudio, head over to “My Models” and click to open up the configuration settings interface for the model.

You’ll get a chance to view all the knobs that AI models have. And as I mentioned, there are a lot of them.

But the one we care about right now is the Context Length. We can see that the default length for this model is 4096 tokens. But it supports up to 8192 tokens. Let’s max it out!

LMStudio provides a helpful warning and probable reason for why the model doesn’t default to the max. The context length takes memory and resources. And raising it to “a high value” can impact performance and usage. So if this model had a max length of 40,960 tokens (the Qwen3 model I use sometimes has that high of a max), you might not want to just max it out right away. Instead, increase it by a little at a time to find the sweet spot: a context length big enough for the job, but not oversized.

As network engineers, we are used to fine-tuning knobs for timers, frame sizes, and so many other things. This is right up our alley!

Once you’ve updated your context length, you’ll need to “Eject” and “Reload” the model for the setting to take effect. But once that’s done, it’s time to take advantage of the change we’ve made!

And look at that, with the larger context window, the AI assistant was able to go through the logs and give us a nice write-up about what they show.

I particularly like the shade it threw my way: “…consider seeking assistance from … a qualified network engineer.” Well played, AI. Well played.

But bruised ego aside, we can continue the AI assisted troubleshooting with something like this.

And we’re off to the races. We’ve been able to leverage our AI assistant to:

1. Process a significant amount of log and debug data to identify possible issues
2. Develop a timeline of the problem (that will be super useful in the help desk ticket and root cause analysis documents)
3. Identify some next steps we can do in our troubleshooting efforts.

All stories must end…

And so you have it, our first AI Nerd Knob—Context Length. Let’s review what we learned:

1. AI models have a “working memory” that is referred to as “context length.”
2. Context Length is measured in “tokens.”
3. Oftentimes times an AI model will support a higher context length than the default setting.
4. Increasing the context length will require more resources, so make changes slowly, don’t just max it out completely.

Now, depending on what AI tool you’re using, you may NOT be able to adjust the context length. If you’re using a public AI like ChatGPT, Gemini, or Claude, the context length will depend on the subscription and models you have access to. However, there most definitely IS a context length that will factor into how much “working memory” the AI tool has. And being aware of that fact, and its impact on how you can use AI, is important. Even if the knob in question is behind a lock and key. 🙂

If you enjoyed this look under the hood of AI and would like to learn about more options, please let me know in the comments: Do you have a favorite “knob” you like to turn? Share it with all of us. Until next time!

PS… If you’d like to learn more about using LMStudio, my buddy Jason Belk put a free tutorial together called Run Your Own LLM Locally For Free and with Ease that can get you started very quickly. Check it out!

 

Sign up for Cisco U. | Join the  Cisco Learning Network today for free.

Learn with Cisco

X | Threads | Facebook | LinkedIn | Instagram | YouTube

Use  #CiscoU and #CiscoCert to join the conversation.

Read next:

Creating a NetAI Playground for Agentic AI Experimentation

Take an AI Break and Let the Agent Heal the Network

Share:

This piece also appeared in NPR’s Body Electric newsletter. Sign up for the newsletter so you don’t miss the next one.

Let’s talk about movement, not as exercise, but as a secret weapon for creative people.

On a memorable walk with bestselling author Kelly Corrigan, she told me how she’s become a master of getting unstuck: She’s found that simple, physical breaks — like walking, dancing, or just moving around — can reset her mood and focus.

So I wanted to share a few of her tricks, especially for anyone who writes, thinks or stares at a screen for a living (hi, me too)!

Feeling stuck? Hit play and dance it out
Kelly’s favorite fix for writer’s block: a three-minute dance break.

“I love music so much,” she told me. “It’s the cheapest way to change your mood.”

Forget toxic productivity. When the vibe turns sour, don’t power through — shake it off. Literally. Grab your phone, play a song you love, and move. You might be one verse away from your next great idea.

Get up while you’re catching up
Writing can feel like a sedentary sport, but Kelly tries to get upright whenever she can, especially during phone calls or research listening sessions.

Not every task allows for it, but many do. Scheduling a dentist appointment? Returning a call? Try doing it on foot.

Make your break…beautiful
When Kelly finishes a movement break, she’ll often take a minute to tidy her space, clear her desk, take out the recycling, wipe down a counter. It’s not about perfection. It’s about caring for the environment where her ideas take shape.

Work with your body — not against it
The biggest insight? Stop trying to be productive all day long. Kelly doesn’t schedule much for early afternoons — she knows that’s her low-energy time. “I don’t fight with myself anymore about how I work and what my periods of rest look like,” she said.

Most of us don’t have a writer’s flexibility, but we can get curious about our own rhythms and make micro-adjustments.

Think of these breaks as just part of the creative process.

This piece was written by Manoush Zomorodi and Fiona Geiran. It was edited by Sanaz Meshkinpour.

Click here to listen to my conversation with Kelly and definitely check out her show Tell Me More on PBS.

Listen to Body Electric on Apple Podcasts and Spotify.