# Research

Studies of coding agents on real C and C++ code: what they get right, what they cost, and how we measured it.

1. **[ByteAsk: Build-Aware Context and a Checking Harness for C++ Coding Agents](https://byteask.ai/research/byteask-system-cpp):** Nov 9, 2026 Paper · Preprint ByteAsk is a C++ coding agent that uses the project's build settings to find the relevant code, gives the AI model that context before it starts, and builds and runs only the tests a change reaches. On the same AI model as Codex CLI, with the same number of runs fixed, it used 63–80% fewer tokens and cost 59–70% less. Pratyush Saini, CTO, ByteAsk Research; Anirudha Kulkarni, CEO, ByteAsk Research; Aman Gupta, Carnegie Mellon University [Read the paper](https://byteask.ai/research/byteask-system-cpp) [PDF, 59 KB](https://byteask.ai/papers/byteask-system-cpp-2026-11.pdf?v=11) [Code](https://github.com/ByteAsk/byteask-system-cpp)
2. **[Beyond Text Search: Context Engines for C++](https://byteask.ai/research/context-engines-cpp):** Nov 9, 2026 Paper · Preprint ByteAsk Context is a context engine for C and C++ coding agents that understands a codebase the way the project's own build does. On real issues from open-source C and C++ projects, it put a file that needed changing first about twice as often as sgrep and almost three times as often as Probe. Pratyush Saini, CTO, ByteAsk Research; Anirudha Kulkarni, CEO, ByteAsk Research; Aman Gupta, Carnegie Mellon University [Read the paper](https://byteask.ai/research/context-engines-cpp) [PDF, 57 KB](https://byteask.ai/papers/context-engines-cpp-2026-11.pdf?v=9) [Code](https://github.com/ByteAsk/context-engines-cpp)
3. **[Build-Aware Context for Large C++ Codebases](https://byteask.ai/research/context-large-codebases-cpp):** Nov 9, 2026 Paper · Preprint ByteAsk Context understands C and C++ code the way the project's build does and tells a coding agent where a change belongs. On one machine it indexed the Linux kernel's build in 9.3 minutes and LLVM's in 13.7, re-indexed a source-file edit in 5.8 seconds, and ranked a changed file first on real open-source issues about twice as often as sgrep. Pratyush Saini, CTO, ByteAsk Research; Anirudha Kulkarni, CEO, ByteAsk Research; Aman Gupta, Carnegie Mellon University [Read the paper](https://byteask.ai/research/context-large-codebases-cpp) [PDF, 59 KB](https://byteask.ai/papers/context-large-codebases-cpp-2026-11.pdf?v=4) [Code](https://github.com/ByteAsk/context-large-codebases-cpp)
4. **[Claude Code with a Language Server: Higher Cost, No Clear Gain in C/C++ Bug Fixes](https://byteask.ai/research/lsp-claude-code-cpp):** Comparing Claude Code with and without the clangd code navigation tool on 255 issues from 40 software projects. Nov 4, 2026 Paper · Preprint We tested whether a tool that helps Claude Code find its way around code would improve bug fixes. Across 255 C and C++ issues, the tested setup cost 11% more overall, with no clear improvement in successful fixes. Pratyush Saini, CTO, ByteAsk Research; Anirudha Kulkarni, CEO, ByteAsk Research [Read the paper](https://byteask.ai/research/lsp-claude-code-cpp) [PDF, 213 KB](https://byteask.ai/papers/why-lsp-claude-code-cpp-2026-11.pdf) [Code](https://github.com/ByteAsk/lsp-claude-code-cpp)

---
HTML version: https://byteask.ai/research