Triboelectric nanogenerators (TENGs) have attracted increasing interest in self-powered sensing and wearable energy harvesting; however, the role of electroactive phase engineering in fully flexible, fiber-based TENGs remains insufficiently explored. Here, we proposed a flexible, all-fiber electroactive-phase-assisted TENG (AF-TENG) comprising BaTiO₃ nanoparticles (BTO NPs)-incorporated polyvinylidene difluoride (PVDF) nanofibers as the tribo-negative layer and thermoplastic polyurethane (TPU) nanofibers as the tribo-positive layer. By synergistically combining electrospinning-induced in situ poling, mechanical stretching, and BTO-induced β-phase nucleation, the electroactive phase of PVDF was effectively tuned at the material level. FTIR analysis quantified the electroactive phase through the β-phase fraction of 71% for PVDF-1%BTO compared to 63% for pure PVDF. As a result, the optimized AF-TENG delivered an output voltage of ~73   V and a current of ~7.7   µA at 20   N and 4   Hz, whereas a peak power of ~55.68   µW and a power density of ~6.18   µW   cm⁻² were observed. Unlike previously reported PVDF-BTO-based TENGs that rely on rigid or semi-rigid counter layers, the fully fibrous architecture presented here offers superior flexibility, conformability, and wearability, which makes it particularly suitable for on-body applications. The AF-TENG demonstrated practical functionality by charging capacitors, lighting LEDs, and powering a commercial stopwatch. Moreover, by integrating machine learning algorithms, the device enabled highly accurate human joint angle prediction, establishing a novel framework in which an all-fiber TENG simultaneously functions as an energy harvester and an intelligent, self-powered motion sensor. This work highlights electroactive phase engineering in an all-fiber platform as an effective strategy for advancing wearable, and intelligent self-powered sensing systems.