Issues

• swe-router: calibrate the quality floors and check the ranking on a second repository

  enhancement
  aarora79/agentic-coding-harness-benchmarks#221 · Filter by author aarora79 opened on Oct 11, 2026
  1

• Flaky test: test_untracked_file_written_during_the_level_is_quarantined fails about 1 in 15 runs

  bug
  aarora79/agentic-coding-harness-benchmarks#211 · Filter by author aarora79 opened on Oct 5, 2026

• Benchmark the p5en open-weight models under the Claude Code harness on mcp-gateway-registry-v2

  enhancement
  aarora79/agentic-coding-harness-benchmarks#155 · Filter by author aarora79 opened on Sep 1, 2026

• Benchmark the g6e open-weight models under the Claude Code harness on mcp-gateway-registry-v2

  enhancement
  aarora79/agentic-coding-harness-benchmarks#154 · Filter by author aarora79 opened on Sep 1, 2026

• Benchmark the Anthropic models under the Claude Code harness on mcp-gateway-registry-v2

  enhancement
  aarora79/agentic-coding-harness-benchmarks#153 · Filter by author aarora79 opened on Sep 1, 2026

• Benchmark the p5en.48xlarge model set under the opencode harness (glm-5.2, kimi-k2.7-code, minimax-m2.5, qwen3-coder-next, nemotron-ultra-550b)

  enhancement
  aarora79/agentic-coding-harness-benchmarks#75 · Filter by author aarora79 opened on Aug 1, 2026
  1

• Benchmark the p5en.48xlarge model set under the pi harness (glm-5.2, kimi-k2.7-code, minimax-m2.5, qwen3-coder-next, nemotron-ultra-550b)

  enhancement
  aarora79/agentic-coding-harness-benchmarks#74 · Filter by author aarora79 opened on Aug 1, 2026
  1

• End-to-end benchmark harness support for opencode with vLLM (like --agent pi)

  enhancementhelp wanted
  aarora79/agentic-coding-harness-benchmarks#72 · Filter by author aarora79 opened on Aug 1, 2026
  2