I used the series expansion of $e^{-x}$ and the fact that $\log(x)$ was less than $x$ in the $(0, \infty)$ to get an upper bound and so use simple comparison to show this was indeed integrable over $(0,\infty)$.
Can I do any better? Definitely can't integrate by parts or use any familiar tools but can I write in the form of a summation or something? According to Wolfram the answer is $\gamma$ the E-M constant.
Edit: Nope that's a horrific mistake: "$\log(x)$ was less than $x$"
Edit: I have realised what I said works in $(1,\infty)$. The problem is $(0,1)$ On $(1,\infty)$ $log(x) < x$ and $e^{x}=1+x+...+\frac{x^3}{6}...\geq \frac{x^3}{6}$ which is integrable.
so $e^{-x}\log(x) \leq \frac{6}{x^2} $
we integrate that in $(1,\infty)$ to get $6$.
The other region I can't figure out
You should break apart the interval and isolate positive and negative contributions. The correct value is $- \gamma$ not $\gamma$.
On $[0,1]$ we have $e^{-1} \leqslant e^{-x} \leqslant 1$ and $\log x \leqslant \log x e^{-x} \leqslant e^{-1} \log x.$
Hence,
$$\int_0^1 \log x \, dx = -1 \leqslant \int_0^1 \log x e^{-x} \, dx \leqslant - e^{-1}.$$
On $[1 , \infty)$ we have using $\log x = 2 \log \sqrt{x} < 2(\sqrt{x} -1)$
$$0 < \int_1^\infty \log x e^{-x} \, dx \leqslant \int_1^\infty 2(\sqrt{x}-1) e^{-x} dx = \sqrt{\pi} \text{erfc}(1) =0.278806.$$
So a first cut (that beats $6$ and at least shows that the value is negative) is
$$-1 \leqslant \int_0^\infty \log x e^{-x} dx \leqslant -0.089073$$
Using the less tight bound $\log x < x -1$ we get $0$ as the upper bound for the integral.