= Solution
Work on the physical domain
$$
s=S+\bar S>0,\qquad t=T+\bar T-|C|^2>0,
$$
and abbreviate $A=ae^{-\alpha S}$, $W=C^3+A+b$, and $B=W-sW_S=C^3+b+(1+\alpha s)A$. There are no <vector multiplets> specified, hence no additional gauge <D-term>. In Planck units the <supergravity F-term potential> and upper-index <supergravity auxiliary fields> are
$$
V=e^K(K^{i\bar j}D_iW\,\overline{D_jW}-3|W|^2),\qquad
F^i=-e^{K/2}K^{i\bar j}\overline{D_jW},\qquad e^K=\frac1{st^3}.
$$
The overall auxiliary-field sign is a convention; its vanishing and <norm> are invariant statements.
The <Kähler covariant derivatives of a superpotential> here are
$$
D_SW=-\alpha A-\frac Ws=-\frac Bs,\qquad
D_TW=-\frac{3W}t,\qquad
D_CW=3C^2+\frac{3\bar C W}t.
$$
The <Kähler metric>, with rows $S,T,C$ and columns their barred partners, is
$$
K_{i\bar j}=\begin{pmatrix}
s^{-2}&0&0\\
0&3/t^2&-3C/t^2\\
0&-3\bar C/t^2&3/t+3|C|^2/t^2
\end{pmatrix}.
$$
Its $T,C$ block is positive definite because its leading diagonal and determinant $9/t^3$ are positive. In the inverse-tensor convention $K_{i\bar j}K^{k\bar j}=\delta_i^k$, the nonzero upper components are
$$
K^{S\bar S}=s^2,\qquad
\begin{pmatrix}K^{T\bar T}&K^{T\bar C}\\K^{C\bar T}&K^{C\bar C}\end{pmatrix}
=\frac t3\begin{pmatrix}t+|C|^2&\bar C\\C&1\end{pmatrix}.
$$
In particular the conjugations in the off-diagonal components matter.
To verify the <no-scale supergravity> cancellation explicitly, temporarily write $u=D_TW=-3W/t$ and $v=D_CW=W_C+3\bar CW/t$. Their part of the metric contraction is
$$
\frac t3\{(t+|C|^2)|u|^2+\bar C\,u\bar v+C\,v\bar u+|v|^2\}
=3|W|^2+\frac t3|W_C|^2.
$$
The terms involving $C W_C\bar W$ and its conjugate cancel between the mixed and diagonal contributions. The remaining $3|W|^2$ cancels the universal negative term. Since $W_C=3C^2$, the exact <matter-logarithm no-scale potential> is therefore
$$
\boxed{V=\frac{|C^3+b+(1+\alpha s)ae^{-\alpha S}|^2}{st^3}
+\frac{3|C|^4}{st^2}\ge0.}
$$
It differs from a model where $|C|^2$ appears outside the logarithm in the <Kähler potential>.
Substitution into the auxiliary-field formula gives
$$
\boxed{F^S=e^{K/2}s\bar B,\qquad
F^T=e^{K/2}t(\bar A+\bar b),\qquad
F^C=-e^{K/2}t\bar C^2.}
$$
For example $F^T=e^{K/2}t(\bar W-\bar C\,\overline{W_C}/3)$, which reduces to the displayed expression for the cubic <superpotential>. To test unbroken <supersymmetry> at any finite point, $F^C=0$ first forces $C=0$. Then $F^T=0$ forces $A+b=0$, whereas $F^S=0$ forces $(1+\alpha s)A+b=0$. Subtracting gives $\alpha sA=0$. Because $s>0$, $\alpha>0$ and $e^{-\alpha S}\ne0$ at a finite point, all auxiliaries can vanish only if \b[$a=b=0$]. Thus every nontrivial choice of $(a,b)$ breaks <supersymmetry> at finite field values.
A finite zero-energy minimum requires both nonnegative terms of $V$ to vanish:
$$
\boxed{C=0,\qquad b+(1+\alpha s)ae^{-\alpha S}=0.}
$$
There cannot be a finite positive-energy stationary point: keeping $S,C$ fixed and varying the real part of $T$, equivalently $t$, gives
$$
\frac{\partial V}{\partial t}=-\frac{3|B|^2}{st^4}-\frac{6|C|^4}{st^3},
$$
which is strictly negative unless $B=C=0$. Hence these zero-energy solutions are all the possible finite minima.
Because $a,b$ were allowed to be arbitrary, existence must be checked. If both are nonzero, put $x=\alpha s>0$ and $r=|b/a|$. The magnitude equation is
$$
r=(1+x)e^{-x/2}.
$$
The function on the right starts at one at the excluded boundary $x=0$, tends to zero as $x\to\infty$, and has derivative $\tfrac12(1-x)e^{-x/2}$. Its maximum is $2e^{-1/2}$ at $x=1$. Therefore a finite <cubic-matter no-scale vacuum with an exponential dilaton> exists exactly when
$$
\boxed{a,b\ne0\quad\text{and}\quad0<|b/a|\le2e^{-1/2}.}
$$
There is one finite positive root for $0<r\le1$, two for $1<r<2e^{-1/2}$, and one merged root at the maximum. The complex phase equation fixes $\operatorname{Im}S$ modulo $2\pi/\alpha$. If exactly one of $a,b$ vanishes, or the ratio exceeds the bound, no finite minimum exists; the potential instead has infimum zero along the runaway $t\to\infty$. A zero infimum at an infinite-field boundary is not an attained finite vacuum.
On a nontrivial finite branch, $W=A+b=-\alpha sA\ne0$, so
$$
\boxed{V_{\min}=0,\qquad F^S=F^C=0,\qquad
F^T=e^{K/2}t\bar W\ne0.}
$$
This proves <supersymmetry breaking> at the intended no-scale vacuum. Its zero energy is consistent with the negative gravitational term: $K_{T\bar T}|F^T|^2=3e^K|W|^2$, which cancels $3e^K|W|^2$ in $V$. It is not a counterexample to the global-supersymmetry positivity argument in question 3.
Both real components of $T$ are exactly flat on the vacuum family: $\operatorname{Im}T$ never enters the potential, and at $B=C=0$ every allowed positive $\operatorname{Re}T$ gives the same zero energy. Generically both components of $S$ are fixed. Although the matter field has no quadratic mass there, it is not an exact flat direction: for $S$ fixed at its root, $B=C^3$ and the leading positive potential is $3|C|^4/(st^2)$. At the special merged root $x=1$, the real-$S$ quadratic curvature also vanishes, but higher orders lift it; this does not add a continuous flat vacuum direction. If $a=b=0$, by contrast, $C=0$ is a supersymmetric zero-energy family with both complex $S$ and $T$ flat and all auxiliaries zero. This exception prevents an unconditional breaking assertion for arbitrary constants.
Finally the <gravitino mass from a superpotential> is
$$
\boxed{m_{3/2}=e^{K/2}|W|=\frac{|C^3+ae^{-\alpha S}+b|}{\sqrt{st^3}}.}
$$
At a nontrivial finite minimum it becomes
$$
\boxed{m_{3/2}=\frac{\alpha\sqrt s\,|ae^{-\alpha S}|}{t^{3/2}}
=\frac{\alpha\sqrt s\,|b|}{(1+\alpha s)t^{3/2}}>0.}
$$
Its value varies along the flat volume modulus; the classical potential does not select a unique supersymmetry-breaking or gravitino-mass scale. On the trivial supersymmetric branch $a=b=C=0$, it is zero.
Back to article page