The Mannequin Context Protocol (MCP) is genuinely helpful. It provides individuals who develop AI instruments a standardized solution to name features and entry information from exterior techniques. As an alternative of constructing customized integrations for every information supply, you possibly can expose databases, APIs, and inside instruments by way of a typical protocol that any AI can perceive.
Nonetheless, I’ve been watching groups undertake MCP over the previous yr, and I’m seeing a disturbing sample. Builders are utilizing MCP to shortly join their AI assistants to each information supply they’ll discover—buyer databases, help tickets, inside APIs, doc shops—and dumping all of it into the AI’s context. And since the AI is sensible sufficient to kind by way of an enormous blob of knowledge and pick the elements which might be related, all of it simply works! Which, counterintuitively, is definitely an issue. The AI cheerfully processes large quantities of knowledge and produces cheap solutions, so no person even thinks to query the strategy.
That is information hoarding. And like bodily hoarders who can’t throw something away till their houses turn into so cluttered they’re unliveable, information hoarding has the potential to trigger critical issues for our groups. Builders be taught they’ll fetch way more information than the AI wants and supply it with little planning or construction, and the AI is sensible sufficient to cope with it and nonetheless give good outcomes.
When connecting a brand new information supply takes hours as a substitute of days, many builders don’t take the time to ask what information really belongs within the context. That’s how you find yourself with techniques which might be costly to run and inconceivable to debug, whereas a complete cohort of builders misses the possibility to be taught the essential information structure expertise they should construct sturdy and maintainable purposes.
How Groups Study to Hoard
Anthropic launched MCP in late 2024 to provide builders a common solution to join AI assistants to their information. As an alternative of sustaining separate code for connectors to let AI entry information from, say, S3, OneDrive, Jira, ServiceNow, and your inside DBs and APIs, you utilize the identical easy protocol to offer the AI with all kinds of knowledge to incorporate in its context. It shortly gained traction. Corporations like Block and Apollo adopted it, and groups in all places began utilizing it. The promise is actual; in lots of instances, the work of connecting information sources to AI brokers that used to take weeks can now take minutes. However that velocity can come at a value.
Let’s begin with an instance: a small group engaged on an AI software that reads buyer help tickets, categorizes them by urgency, suggests responses, and routes them to the proper division. They wanted to get one thing working shortly however confronted a problem: They’d buyer information unfold throughout a number of techniques. After spending a morning arguing about what information to drag, which fields had been needed, and learn how to construction the combination, one developer determined to simply construct it, making a single getCustomerData(customerId) MCP software that pulls the whole lot they’d mentioned—40 fields from three completely different techniques—into one massive response object. To the group’s reduction, it labored! The AI fortunately consumed all 40 fields and began answering questions, and no extra discussions or choices had been wanted. The AI dealt with all the brand new information simply positive, and everybody felt just like the venture was heading in the right direction.
Day two, somebody added order historical past so the assistant may clarify refunds. Quickly the software pulled Zendesk standing, CRM standing, eligibility flags that contradicted one another, three completely different identify fields, 4 timestamps for “final seen,” plus total dialog threads, and mixed all of them into an ever-growing information object.
The assistant saved producing reasonable-looking solutions, at the same time as the information it ingested saved rising in scale. Nonetheless, the mannequin now needed to wade by way of 1000’s of irrelevant tokens earlier than answering easy questions like “Is that this buyer eligible for a refund?” The group ended up with a knowledge structure that buried the sign in noise. That extra load put stress on the AI to dig out that sign, resulting in critical potential long-term issues. However they didn’t understand it but, as a result of the AI saved producing reasonable-looking solutions. As they added extra information sources over the next weeks, the AI began taking longer to reply. Hallucinations crept in that they couldn’t monitor all the way down to any particular information supply. What had been a very worthwhile software grew to become a bear to take care of.
The group had fallen into the information hoarding entice: Their early fast wins created a tradition the place folks simply threw no matter they wanted into the context, and finally it grew right into a upkeep nightmare that solely obtained worse as they added extra information sources.
The Abilities That By no means Develop
There are as many opinions on information structure as there are builders, and there are often some ways to unravel anybody drawback. One factor that just about everybody agrees on is that it takes cautious decisions and plenty of expertise. Nevertheless it’s additionally the topic of plenty of debate, particularly inside groups, exactly as a result of there are such a lot of methods to design how your utility shops, transmits, encodes, and makes use of information.
Most of us fall into just-in-case considering at one time or one other, particularly early in our careers—pulling all the information we would presumably want simply in case we’d like it quite than fetching solely what we’d like once we really want it (which is an instance of the alternative, just-in-time considering). Usually once we’re designing our information structure, we’re coping with rapid constraints: ease of entry, measurement, indexing, efficiency, community latency, and reminiscence utilization. However once we use MCP to offer information to an AI, we will typically sidestep a lot of these trade-offs…quickly.
The extra we work with information, the higher we get at designing how our apps use it. The extra early-career builders are uncovered to it, the extra they be taught by way of expertise why, for instance, System A ought to personal buyer standing whereas System B owns fee historical past. Wholesome debate is a crucial a part of this studying course of. Via all of those experiences, we develop an instinct for what “an excessive amount of information” appears to be like like—and learn how to deal with all of these tough however essential trade-offs that create friction all through our tasks.
MCP can take away the friction that comes from these trade-offs by letting us keep away from having to make these choices in any respect. If a developer can wire up the whole lot in only a few minutes, there’s no want for dialogue or debate about what’s really wanted. The AI appears to deal with no matter information you throw at it, so the code ships with out anybody questioning the design.
With out all of that have making, discussing, and debating information design decisions, builders miss the possibility to construct essential psychological fashions about information possession, system boundaries, and the price of transferring pointless information round. They spend their youth connecting as a substitute of architecting. That is one other instance of what I name the cognitive shortcut paradox—AI instruments that make improvement simpler can forestall builders from constructing the very expertise they should use these instruments successfully. Builders who rely solely on MCP to deal with messy information by no means be taught to acknowledge when information structure is problematic, similar to builders who rely solely on instruments like Copilot or Claude Code to generate code by no means be taught to debug what it creates.
The Hidden Prices of Information Hoarding
Groups use MCP as a result of it really works. Many groups fastidiously plan their MCP information structure, and even groups that do fall into the information hoarding entice nonetheless ship profitable merchandise. However MCP remains to be comparatively new, and the hidden prices of knowledge hoarding take time to floor.
Groups typically don’t uncover the issues with a knowledge hoarding strategy till they should scale their purposes. That bloated context that hardly registered as a value on your first hundred queries begins displaying up as an actual line merchandise in your cloud invoice whenever you’re dealing with thousands and thousands of requests. Each pointless discipline you’re passing to the AI provides up, and also you’re paying for all that redundant information on each single AI name.
Any developer who’s handled tightly coupled courses is aware of that when one thing goes unsuitable—and it all the time does, finally—it’s so much more durable to debug. You typically find yourself coping with shotgun surgical procedure, that actually disagreeable scenario the place fixing one small drawback requires adjustments that cascade throughout a number of elements of your codebase. Hoarded information creates the identical form of technical debt in your AI techniques: When the AI provides a unsuitable reply, monitoring down which discipline it used or why it trusted one system over one other is tough, typically inconceivable.
There’s additionally a safety dimension to information hoarding that groups typically miss. Every bit of knowledge you expose by way of an MCP software is a possible vulnerability. If an attacker finds an unprotected endpoint, they’ll pull the whole lot that software offers. If you happen to’re hoarding information, that’s your total buyer database as a substitute of simply the three fields really wanted for the duty. Groups that fall into the information hoarding entice discover themselves violating the precept of least privilege: Functions ought to have entry to the information they want, however no extra. That may convey an unlimited safety threat to their entire group.
In an excessive case of knowledge hoarding infecting a complete firm, you may uncover that each group in your group is constructing their very own blob. Assist has one model of buyer information, gross sales has one other, product has a 3rd. The identical buyer appears to be like utterly completely different relying on which AI assistant you ask. New groups come alongside, see what seems to be working, and replica the sample. Now you’ve obtained information hoarding as organizational tradition.
Every group thought they had been being pragmatic, transport quick, and avoiding pointless arguments about information structure. However the hoarding sample spreads by way of a company the identical means technical debt spreads by way of a codebase. It begins small and manageable. Earlier than it, it’s in all places.
Sensible Instruments for Avoiding the Information Hoarding Entice
It may be actually tough to teach a group away from information hoarding after they’ve by no means skilled the issues it causes. Builders are very sensible—they wish to see proof of issues and aren’t going to take a seat by way of summary discussions about information possession and system boundaries when the whole lot they’ve carried out thus far has labored simply positive.
In Studying Agile, Jennifer Greene and I wrote about how groups resist change as a result of they know that what they’re doing at the moment works. To the particular person attempting to get builders to alter, it could seem to be irrational resistance, however it’s really fairly rational to push again towards somebody from the skin telling them to throw out what works at the moment for one thing unproven. However similar to builders finally be taught that taking time for refactoring speeds them up in the long term, groups must be taught the identical lesson about deliberate information design of their MCP instruments.
Listed here are some practices that may make these discussions simpler, by beginning with constraints that even skeptical builders can see the worth in:
- Construct instruments round verbs, not nouns. Create
checkEligibility()orgetRecentTickets()as a substitute ofgetCustomer(). Verbs drive you to consider particular actions and naturally restrict scope. - Discuss minimizing information wants. Earlier than anybody creates an MCP software, have a dialogue about what the smallest piece of knowledge they should present for the AI to do its job is and what experiments they’ll run to determine what the AI really wants.
- Break reads other than reasoning. Separate information fetching from decision-making whenever you design your MCP instruments. A easy
findCustomerId()software that returns simply an ID makes use of minimal tokens—and won’t even must be an MCP software in any respect, if a easy API name will do. ThengetCustomerDetailsForRefund(id)pulls solely the particular fields wanted for that call. This sample retains context centered and makes it apparent when somebody’s attempting to fetch the whole lot. - Dashboard the waste. The perfect argument towards information hoarding is displaying the waste. Monitor the ratio of tokens fetched versus tokens used and show them in an “data radiator” fashion dashboard that everybody can see. When a software pulls 5,000 tokens however the AI solely references 200 in its reply, everybody can see the issue. As soon as builders see they’re paying for tokens they by no means use, they get very occupied with fixing it.
Fast scent check for information hoarding
- Instrument names are nouns
(getCustomer())as a substitute of verbs(checkEligibility()). - No one’s ever requested, “Do we actually want all these fields?”
- You may’t inform which system owns which piece of knowledge.
- Debugging requires detective work throughout a number of information sources.
- Your group not often or by no means discusses the information design of MCP instruments earlier than constructing them.
Wanting Ahead
MCP is an easy however highly effective software with monumental potential for groups. However as a result of it may be a critically vital pillar of your total utility structure, issues you introduce on the MCP stage ripple all through your venture. Small errors have big penalties down the highway.
The very simplicity of MCP encourages information hoarding. It’s a straightforward entice to fall into, even for knowledgeable builders. However what worries me most is that builders studying with these instruments proper now may by no means be taught why information hoarding is an issue, and so they received’t develop the architectural judgment that comes from having to make onerous decisions about information boundaries. Our job, particularly as leaders and senior engineers, is to assist everybody keep away from the information hoarding entice.
If you deal with MCP choices with the identical care you give any core interface—maintaining context lean, setting boundaries, revisiting them as you be taught—MCP stays what it needs to be: a easy, dependable bridge between your AI and the techniques that energy it.
