We propose to use the last Spitzer observing cycle to increase the sample of hidden post-AGB stars with mid-infrared spectroscopy. We selected predominantly GLIMPSE selected obscured objects tentatively identified as stars which have just abandoned the Asymptotic Giant Branch (AGB) and are now evolving to the planetary nebula (PN) stage. They are expected to represent the higher mass fraction of former oxygen-rich AGB stars and are suspected to be precursors of type I PNe. A high fraction of them shows a near infrared excess discovered in the GLIMPSE bands, indicating that as a consequence of strongly decreasing mass loss rates the surrounding circumstellar dust shell is becoming transparent again. Our sample is relatively faint at IRAS wavelengths so that they do not appear in IRAS color-selected samples. Based on our experience with Spitzer observations of an IRAS color selected sample, we will be able 1) to determine/confirm the chemical composition of the dust as O-rich, and possibly identify new mixed chemistry sources 2) to determine the dust grain structure (amorphous vs. crystalline) 3) to identify the solid state features superimposed on the dust continuum, in particular the relative amount of Al, Mg, Fe in the dust crystalline silicates 4) to study the correlation of the above observational properties with the evolutionary stage of the source, and 5) to identify new young infrared PNe, which may be associated to the high mass population of PNe The proposed Spitzer observing program is part of our efforts to understand the transition from AGB to PN while the stars are still in the earliest stages of the post-AGB phase. We found an astonishingly wide diversity of dust properties in their shells, probably because we observe them at different stages of their fast evolution. Due to their faintness at wavelengths shorter than 3 micron, Spitzer observations will be the last chance to study dust shells of hidden post-AGB stars for a long time.