Agent Friendly Code
← back to leaderboard

ggml-org/llama.cppGitHub

https://github.com/ggml-org/llama.cpp
Stars:
105.9k
Default branch:
master
Last scored:
3h ago
Overall

Strengths

  • AGENTS.md / CLAUDE.md
  • CI configuration
  • CONTRIBUTING guide
  • Dependency manifest
  • Reproducible dev env

Gaps

  • No missing signals — nice.

Suggestions to improve for a specific model

  1. 1
    Manageable size
    If possible, split into smaller modules or carve out a focused entry path. Document where to start in AGENTS.md.
    +3.3 pts

Per-model scores

Claude Code
Weights AGENTS.md and tests heavily — Claude Code leans on an instructions file and a fast feedback loop.
96.7
Cursor
Weights type config and a detailed README highly — Cursor's inline edits benefit from static types and skim-readable docs.
97.2
Devin
Weights CI and reproducible envs highly — Devin runs in a sandboxed VM and needs end-to-end automation.
96.3
GPT-5 Codex
Balanced profile as a reference point.
96.6

Signal breakdown

AGENTS.md / CLAUDE.md· 100% pass
Substantive (6856 chars)
AGENTS.md
CI configuration· 100% pass
35 GitHub Actions workflow(s)
.github/workflows
CONTRIBUTING guide· 100% pass
Guide present
CONTRIBUTING.md
Dependency manifest· 100% pass
Manifest present
pyproject.toml
Reproducible dev env· 100% pass
3 env artifacts (Makefile, makefile)
Makefile
License file· 100% pass
License present
LICENSE
Linter / formatter config· 100% pass
Config detected
.flake8
Pre-commit / git hooks· 100% pass
Hook framework configured
.pre-commit-config.yaml
README· 100% pass
README detailed (30159 chars)
README.md
Manageable size· 50% pass
2423 files — large
Test suite· 100% pass
Found /tests
tests
Type configuration· 100% pass
Type config present
mypy.ini