Understanding SCP Object Classes
The SCP Foundation categorizes its anomalies using a system that sounds simple but gets messy fast. If you are new to this, the three main classes are Safe, Euclid, and Keter, but treating them as a straightforward safety ranking is one of the most common mistakes beginners make. These labels describe containment difficulty, not danger level. A Keter-class object can be less lethal than a Safe object that simply refuses to stay still without special protocols. I have spent years tracking how these classifications shift, and honestly, the system is more fluid than the wiki would have you believe. Anomalous objects get reclassified regularly based on new research, protocol improvements, or failures. What was Keter two years ago might be Euclid today if the site found a reliable containment method. Conversely, a Safe object can escalate to Keter overnight if something changes about its behavior that nobody predicted. Safe objects are those where containment procedures, once established, are expected to reliably prevent anomalous effects. The key word is expected. Reliability is the goal, not perfection. I once dealt with an object classified as Safe that had a one in ten thousand chance per activation cycle of triggering a localized reality failure. Statistically negligible, sure, but we lost a researcher because that one in ten thousand happened on a Tuesday. Safe does not mean harmless. It means we understand the anomaly well enough to predict what happens when we interact with it, and we know how to stop the effects if they occur.
Common pitfalls with Safe classification include assuming the object is stable forever. Several documented cases exist where Safe objects underwent behavioral drift that went undetected for months. The Foundation recommends reviewing Safe-class protocols at least annually, even if nothing has changed visibly.
Euclid Class
Euclid covers anomalies that are unpredictable, variable, or require more effort to contain than Safe objects provide. This is the largest category, and it frustrates researchers because it represents genuine uncertainty. You cannot write a protocol that guarantees containment because the object may behave differently next time. I worked on a site that had a Euclid object which responded to observation frequency. The more researchers monitored it, the more aggressively it tried to escape containment. We had to develop a rotating watch schedule that deliberately kept the object under minimal observation while maintaining baseline awareness. The workaround was tedious. We set up automated sensors that triggered only when movement exceeded normal parameters, reduced human oversight to a two-person minimum with staggered shifts, and accepted that containment would never be as tight as we wanted it. That is Euclid work. You optimize, you adapt, and you accept that it might not hold forever.
Get the Full Details
Keter Class
Keter objects resist containment or pose a significant threat even when contained. These are the nightmares of the Foundation. A Keter classification means current protocols are insufficient to guarantee long-term containment, or the consequences of containment failure are severe. I have seen Keter objects that required entire site budgets just to maintain their containment procedures, with constant upgrades and emergency protocols cycling through every few months. One thing the public rarely understands is that Keter is not a life sentence. I was part of a team that spent eighteen months downgrading a Keter-class entity to Euclid. The object had a predictable activation pattern that we finally mapped after tracking thirty-seven separate incidents. Once we understood the pattern, we redesigned the containment chamber around it. The downgrade took nearly two years and cost roughly four million credits in equipment and personnel time, but it worked. That is why the Foundation invests in research teams even for Keter objects. They are not giving up.
Other Classes
Beyond the big three, the Foundation uses several additional classifications that do not get enough attention. Thaumiel objects are anomalies used to contain other anomalies. I handled a Thaumiel that was essentially a modified reality anchor paired with a cognitohazardous containment field. It looked unremarkable until you read the full protocol file. The real value of Thaumiel objects is that they can turn Keter-class problems into Euclid-class problems without triggering the public attention that direct containment would cause. Neutralized means the object is no longer anomalous. Explained means the anomaly has been fully understood and potentially replicated through known physics. Chimerra is for objects that do not fit any existing class and require ad-hoc categorization. Apollyon is reserved for anomalies that cannot be contained and threaten global or universal-scale termination events. I have never personally worked on an Apollyon object, and I prefer to keep it that way. Those sit at O5-level oversight with direct involvement from the highest-clearance researchers in the Foundation.
How Classification Actually Works In Practice
The classification process starts with a proposal from the site where the object was contained, followed by review from the SCP Classification Committee. The committee has five members who vote, and a majority determines the class. Appeals are possible but rare because overturning a classification requires new evidence, not just disagreement. I have seen appeals denied because the committee felt the applicant misunderstood their own data, which happens more often than you might think. The real bottleneck in classification is documentation. An object cannot be properly classified without exhaustive testing records, and testing an unknown anomaly is inherently dangerous. Sites sometimes rush this process under pressure from command, leading to premature classifications that later cause problems. I have personally flagged three instances where an object was classified as Safe before essential behavioral tests were completed. All three turned out to be Euclid-class upon further study. The rule of thumb is simple: do not finalize a classification until you have at least ninety days of consistent observation data, regardless of how confident the site director is.
Common Misconceptions
The biggest misconception is that higher class equals more dangerous. A Safe-class object with a high mortality rate upon containment breach is still Safe if the containment is straightforward and reliable. A Keter-class object that rarely causes casualties can still be Keter if it constantly adapts to escape methods. The class measures containment difficulty, not threat potential. Another misconception is that classification is permanent. It is not. New research, protocol changes, and behavioral shifts all trigger re-evaluation. Some objects have been reclassified three or four times during active containment periods. There is also a misunderstanding about Euclid being a middle ground. It is not. Euclid is essentially the catch-all for everything the Foundation does not fully understand yet. It is the most honest classification the system offers because it admits uncertainty. Safe and Keter both imply a degree of predictability that Euclid explicitly denies.
What This Means For Your Work
If you are reading this because you are involved in containment work or studying anomaly research, the practical takeaway is that classification is a tool, not a verdict. Use it to allocate resources and set expectations, but do not let it make you complacent. A Safe classification should never stop you from reviewing protocols. A Keter classification should never make you stop trying. The object does not care what box you put it in. It will behave according to its nature, and your job is to keep up with it. The Foundation's classification system is imperfect, and everyone who works in it knows it. The alternatives are worse. Without a shared framework for discussing containment difficulty, sites would operate in complete isolation, duplicating efforts and missing patterns that only become visible when data is shared across the organization. The system exists because the alternative is chaos, and chaos is something the Foundation was created to prevent.
