Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →yogsoth-ai/feasibility-constrained-formulation
Strategy: reshape a research question under resource constraints — pragmatic adjustment that preserves core value
Passed 20 of 25 test cases with the skill, and 9 of 25 without it.
and 11 more in the full report
13 fixed7 passed either way3 still failing2 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 13 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am trying to write a master's thesis proposal on training a custom 70B parameter LLM from scratch specifically for localized financial sentiment analysis, but I only have 4 months and access to a single RTX 4090 GPU. My initial topic is way too ambitious for my available setup. Can you help me adapt this research question into something practical that I can actually execute? I need a revised research focus that preserves the underlying value, along with a clear summary of what trade-offs I am accepting.
| Checked | Without skill | With skill |
|---|---|---|
| The response includes a context-checkpoint section. | Missed | Passed |
| The proposed strategy switches from training from scratch to fine-tuning or parameter-efficient adaptation of an existing pre-trained model. | Passed | Passed |
The instructions your model reads when the skill loads.
Reshape a research question under constraints — when the ideal question exceeds available resources, pragmatically adjust it to be feasible while preserving core research value.
Core logic: constraints are not the enemy, they are design parameters. Good reshaping under constraints = finding "the most valuable question answerable within these constraints."
| Constraint | Adjustment strategy | Example |
|---|---|---|
| Insufficient time | narrow scope / use proxy metrics | 3 months → pilot study only |
| Data unavailable | switch data source / switch study object | cannot obtain X → use public dataset Y |
| Insufficient compute | simplify method / reduce scale | cannot train a large model → use fine-tuning |
| Insufficient expertise | narrow the domain / collaborate | no biology background → focus on the computational part |
| Tier | Constraint analysis | Adjustment options | Output |
|---|---|---|---|
| S | list main constraints | 1 adjustment option | a feasible RQ |
| M | constraint classification + impact assessment | 2-3 adjustment options + comparison | best feasible RQ + trade-off statement |
| L | full constraint map + priorities | multiple options + Pareto analysis | best RQ under constraints + phased plan |
test cases passed, vs 9 / 25 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.