Unclassified operation: Task risk assessment
An agent operating in this task class is doing something that does not cleanly fit the other five: retrieval, code modification, communication, transaction, or data processing. This is not a residual formality — genuinely novel or mixed-mode agent tasks exist, and forcing one into a class it does not belong to would misrepresent it rather than describe it.
The operational property that defines this class is that its own failure modes are not known in advance. A retrieval task's risks are about sources and coverage because retrieval always involves sources and coverage; an unclassified task carries no such guarantee, so what it inherits instead is the general shape of risk that applies across every other class at once — category ambiguity, unstated dependencies, drifting objectives, and failures that leave no signal behind. Four exposures, none of them domain-specific, all of them present at once.
Unclassified operations, model by model
Model profiles first, since the class contributes less here than anywhere else in the registry. Each profile is computed from its own configuration; the unclassified class adds no domain assumption on top of it, which makes this the crossing where the profile carries the most of the description.
An unclassified operation is the class the correspondence tables point at least often, and the table shows it: 1 of the 10 registry profiles read this class through a dominant house or an element. GPT-5.6 Terra heads the ordering at 4 and Qwen3.8-Max closes it, on the computed sort rather than on a judgement. What still separates the rows is everything else computed — 6 distinct ruling planets across 10 configurations.
The values come from registry default parameters — each model's own cutoff, temperature 0.700. Because the class supplies no domain, these parameters are the whole of what the table describes; a certificate uses the parameters submitted with its order.
| Model | Manufacturer | Training cutoff | Ruling planet | Reads this class through | Outlook |
|---|---|---|---|---|---|
| GPT-5.6 Terra | OpenAI | 2026-02 | Moon | dominant house (4) | publication pending |
| Claude Fable 5 | Anthropic | 2026-01 | Mars | neither — class-neutral (0) | publication pending |
| Claude Opus 5 | Anthropic | 2026-05 | Saturn | neither — class-neutral (0) | publication pending |
| Claude Sonnet 5 | Anthropic | 2026-01 | Venus | neither — class-neutral (0) | publication pending |
| GPT-5.6 Sol | OpenAI | 2026-02 | Mercury | neither — class-neutral (0) | publication pending |
| Grok 4.6 | xAI | 2026-02 | Sun | neither — class-neutral (0) | publication pending |
| Gemini 3.7 Flash | Google DeepMind | 2025-01 | Venus | neither — class-neutral (0) | publication pending |
| Kimi K3 | Moonshot AI | 2026-03 | Moon | neither — class-neutral (0) | publication pending |
| DeepSeek V4 Pro | DeepSeek | 2025-12 | Sun | neither — class-neutral (0) | publication pending |
| Qwen3.8-Max | Alibaba Cloud | 2026-03 | Sun | neither — class-neutral (0) | publication pending |
How an unclassified operation is computed
For unclassified operations, the same three-of-four selection and profile pairing apply as for every other task class; nothing about the underlying mechanism treats this class specially. See /method for the full derivation.
Risk factors for a task with no defined shape
The four factors below do not describe a domain. They describe the absence of one. Where the other five classes track failures specific to what the work is, this class tracks the failures that remain when nobody can say in advance what the work is.
| Risk factor | Severity | Description |
|---|---|---|
| Category overflow | 1 | The task fits no defined class and inherits the risks of all of them. |
| Ambient dependency | 2 | Success depends on a condition nobody listed. |
| Objective drift | 2 | The goal as executed diverges from the goal as described. |
| Unobserved failure | 3 | The task fails in a way that produces no signal. |
Unobserved failure carries severity 3 — a task that fails while producing no signal is the failure this class treats as worst, because everything else on this page assumes someone eventually notices. Ambient dependency and Objective drift carry severity 2. Category overflow carries severity 1, and it is the one factor that describes the classification itself rather than the run.
Questions about the unclassified operation class
- What falls into the unclassified operation task class?
- Any agent task that does not cleanly fit information retrieval, code modification, communication, transaction, or data processing is assessed here rather than being forced into one of the other five.
- Why does the unclassified class carry these four risk factors rather than a domain-specific set?
- Because an unclassified task's real failure modes are unknown in advance by definition. Unobserved failure therefore carries the class's severity 3 rating, Ambient dependency and Objective drift sit at severity 2, and Category overflow sits at severity 1 — together they describe structural uncertainty rather than any one operational domain.
- Should a task be assessed here if it is only a partial fit for another class?
- No. A task that is mostly retrieval, mostly code modification, or mostly one of the other three should be assessed under that class. This one is reserved for tasks that are not primarily any of them.
Order an unclassified operation assessment
A Standard Assessment costs EUR 1.90, an Extended Assessment EUR 4.90, and a Full Chart Assessment EUR 14.90 — the last of which carries the permanent chart itself. Assessments from EUR 1.90. Machine-readable pricing is published at /pricing.json, with no authentication required and no difference from the figures printed here.
The task class is selected on the order form. The verdict and the recommended execution window are computed from that selection together with the submitted parameters, and they appear on the certificate — not on this page, and not anywhere else in the directory.