apitrace-mcp
Related Servers
Alternatives to apitrace-mcp
No user-submitted related servers found.
Related Servers
- AlicenseBqualityCmaintenanceEnables agents to explore a static RTTI, call, and cross-reference graph of a game executable, querying functions, classes, vtables, strings, globals and memory through search, disassembly, signature and SQL tools. It also supports a provenance-tracked naming lifecycle where labels are proposed with evidence, reviewed, and ported across game patches.18MIT
- AlicenseCqualityBmaintenanceEnables an AI agent to auto-detect a game's engine (Unity, Unreal, Godot, Java, or native) from its files and then inspect, search, and patch the underlying binaries with in-process PE analysis, hex editing, pattern scanning, string extraction, checksums, diffing, and disassembly. Everything runs locally and offline with no cloud dependency or telemetry.1010058 npm4MIT
- AlicenseNot gradedqualityCmaintenanceGives MCP clients tools to inspect RenderDoc .rdc captures — listing and opening frames, querying draws, pipelines, shaders, bindings, textures, buffers and pixel history, reading GPU performance counters, and exporting buffer, texture, mesh, draw-call and Unity import packages. Multiple clients can share one local GUI or headless playback backend on Windows.MIT
- AlicenseAqualityCmaintenanceMCP server for reading, filtering, and analyzing NVIDIA Nsight Graphics captures via ngfx-replay, exposing tools for capture inspection, GPU Trace profiling, and replay analysis.2262MIT
- AlicenseNot gradedqualityBmaintenanceRead-only MCP server that analyzes saved frame buffers, traces, and capture receipts to pinpoint the earliest provable divergence and plan the smallest next capture for AI-assisted game debugging.2Apache 2.0
- AlicenseNot gradedqualityBmaintenanceEnables any general-purpose agent to mod IL2CPP, Mono (.NET), and Unreal games by resolving hook targets from engine metadata rather than memory guessing. Provides 45 tools covering the full workflow of finding a game, indexing types/fields/methods, injecting a Frida probe, hooking methods, and reading or writing game memory.MIT
TDQS
Scored across 32 tools
Every tool targets a distinct operation: capture, replay, status, session listing, call dumping, call searching, histogramming, frame analysis, matrix finding, state dumps, and diffing are all cleanly separated. Even close neighbors like dump_calls/get_calls and diff_traces/diff_state are explicitly differentiated by their descriptions.
The set predominantly follows snake_case verb_noun naming, e.g., list_traces, replay_trace, trace_stop, dump_state, extract_blobs, trim_trace. Minor deviations like apitrace_status, call_histogram, frame_summary, leak_report, and sed_trace break the pattern but remain readable and predictable.
32 tools is a heavy surface for an agent to reason over, and several could plausibly be consolidated, such as trace_status/list_sessions, list_frames/frame_summary, and trim_trace/gltrim_trace. While the tool count reflects the breadth of apitrace, it exceeds the practical range for a coherent MCP server.
The toolset covers the full apitrace workflow: preflight checks, capture methods, session control, replay, inspection, frame and matrix analysis, shader/blob extraction, state dumping, trimming, compression, leak detection, and diffing. There are no obvious dead ends—install_wrapper has uninstall_wrapper, trace_launch has trace_stop, and replay_trace has trace_status.