TargetTalk: a no-code projectile & flight system for Unity. Drag a preset, assign a target — it flies with frame-perfect precision. 84 presets, real-time editor preview, no Play Mode needed.Ships configured for URP. Built-in RP support is included as an optional .unitypackage under "Built-in RP Support/" — import it only if your project uses the Built-in pipeline, then delete the TargetTalk_URP folder. A fresh import leaves no mismatched materials. HDRP is not supported.Design motion with gizmos and values — no code, no animation.TargetTalk (TT) is a Unity system for building projectile and flight choreography without writing code or authoring animation clips. Drag a preset into your scene, shape its trajectory with gizmos, tune the values in the Inspector — that's all it takes.It doesn't replay a fixed path. It talks to its environment.Most motion tools just play back a route you drew in advance. TT's motion holds a real-time conversation with your running game.Automatically responds every frame as the target moves or the distance changesKeeps tracking, waiting, curving, and timing intact — while still honoring its arrival accuracy (the promised hit) to the very endSo the same preset moves differently in every situation, as if it were alive.One system, this much choreographyHovering missiles · throws · magic orbs · arcs · boomerangs · spirals · chains · pets (companions) · flight AI · multi-agent behaviorIt's not a tool for adding more weapon types — it's a tool for designing behavior.Three pillars1. Adaptive MotionEven as a target's position and distance change, TT self-corrects its tracking, curves, and timing. The hit it promises at the moment of firing is kept, no matter how the environment shifts.2. Intuitive DesignShape trajectories with gizmos while watching them with your own eyes, then refine with values. No Play Mode, no compile waits — see results instantly in the editor. Even artists who don't code can finish their motion by hand.3. AI-Assisted AuthoringAn AI Learning Document ships with the package. Feed it to an LLM like ChatGPT or Claude, and it translates natural-language requests — "a projectile that hovers around a flower like a bee, then leaves again" — into TT parameter combinations. A new workflow for designing motion through text.What's included84 preset prefabs (2D · 2.5D · 3D)47 production resources (Trail · Hit · Muzzle)2D / 2.5D / 3D demo scenes included23-language localized InspectorBuilt-in RP · URP both supportedAI Learning Document bundledMotion isn't code — it's a feel in the hands.TargetTalk was built to carry an artist's intent straight into the running game, frame for frame, without a programmer translating it into code. Every parameter, every default, every preset has been hand-tuned.ndledSupported EnvironmentUnity 2022.3 LTS or later (Unity 6 supported)Render pipelines: Built-in RP · URP both supported — shipped as two independent trees; on first import, TT auto-detects your active pipeline and offers to clean up the unused treePlatforms: Standalone (Windows / macOS / Linux), iOS, Android, WebGLNo external package dependencies — import and use right awayHDRP is not supported in the current versionIncluded Content (per pipeline)84 preset prefabs: 2D 27 / 2.5D 27 / 3D 30 (2.5D is a plane-mirror of the 2D behaviors)47 production resources: Trail 29 / Hit 13 / Muzzle 53 demo scenes (2D / 2.5D / 3D)AI Learning Document bundled — for LLM-based motion authoringTechnical FeaturesReal-time editor preview — see and tune trajectories and behavior with no Play ModePer-frame deterministic simulation — no jitter, no driftAuto-responds every frame to target movement and distance changes (tracking, curves, timing, and arrival accuracy all preserved)23-language Inspector localization (tooltips included)AI/ML was used in the following areas:- Localization: the 23-language Inspector localization tables were generated with AI assistance and reviewed by a human.- Documentation: the bundled AI Learning Document was authored with AI assistance. It is meant to be fed to LLMs (ChatGPT, Claude, etc.) so users can author motion via natural language.- Store listing: parts of the store description and its translations were AI-assisted.The core runtime system, simulation algorithm, custom editor tooling, prefabs, parameters, and presets were all designed and hand-tuned by the author — not AI-generated.
