Background: KRAS-G12D mutations drive 20–50% of pancreatic/biliary cancers yet remain challenging to target due to GTP-pocket conservation and high cellular GTP levels. While allosteric inhibitors targeting the SWII pocket (e.g., MRTX1133) show promise, limited chemical diversity and paradoxical cellular/enzymatic activity relationships necessitate the exploration of novel scaffolds. This study aims to develop KRAS-G12D inhibitors and PROTACs to offer a selection of new chemical entities through systematic structure–activity optimization and evaluate their therapeutic potential through PROTAC derivatization. Methods: Eleven compounds featuring heterocyclic cores (pyrimidine/pyrido[4,3-d]pyrimidine/5,6,7,8-tetrahydroprodo[3,4-d]pyrimidine) were designed via structure-based drug design. Antiproliferative activity against KRAS-G12D (Panc1), KRAS-G13D (HCT116) and wild-type (A549) cells was assessed using the CCK-8 assay. KRAS-G12D enzymatic inhibition was measured using a GTPase activity assay. Molecular docking simulations (Sybyl 2.0; PDB:7RPZ) elucidated binding modes. Two PROTACs were synthesized from lead compounds by conjugating E3 ligase linkers. All the novel inhibitors and PROTACs were characterized by means of NMR or HRMS. Results: Compound 10c demonstrated selective anti-proliferation in Panc1 cells (IC50 = 1.40 μM) with 4.9-fold greater selectivity over wild-type cells, despite weak enzymatic inhibition (IC50 > 10 μM). Docking revealed critical hydrogen bonds between its protonated 3,8-diazabicyclo[3.2.1]octane moiety and Asp12/Gly60. The enzymatic inhibitor 10k showed potent KRAS-G12D inhibition (IC50 = 0.009 μM) through homopiperazine-mediated interactions with Glu92/His95. Derived PROTACs 26a/b exhibited reduced potency (IC50 = 3–5 μM vs. parental 10k: 2.22 μM), potentially due to impaired membrane permeability. Conclusions: Eleven novel KRAS-G12D inhibitors with a seven-membered ring pharmacophore were synthesized. Compound 10c showed strong anti-proliferative activity, while 10k exhibited potent enzymatic inhibition. Two PROTACs were designed but showed no clear advantage over 10k. This study provides valuable insights for KRAS-targeted drug development.