Skip to content

Add programmatic tool calling release post with MiniZinc case study#103

Open
r33drichards wants to merge 1 commit intomainfrom
claude/release-post-tool-calling-Puuit
Open

Add programmatic tool calling release post with MiniZinc case study#103
r33drichards wants to merge 1 commit intomainfrom
claude/release-post-tool-calling-Puuit

Conversation

@r33drichards
Copy link
Copy Markdown
Owner

Tutorial and integration test demonstrating mcp-v8's programmatic tool calling feature using the MiniZinc constraint solver MCP server. Covers tool discovery (mcp.servers, mcp.listTools), constraint satisfaction (N-Queens, graph coloring), and optimization (knapsack) — all invoked from JavaScript inside the V8 sandbox via mcp.callTool().

https://claude.ai/code/session_01GZSn58VARmT7FtYkrstnxS

Tutorial and integration test demonstrating mcp-v8's programmatic tool
calling feature using the MiniZinc constraint solver MCP server. Covers
tool discovery (mcp.servers, mcp.listTools), constraint satisfaction
(N-Queens, graph coloring), and optimization (knapsack) — all invoked
from JavaScript inside the V8 sandbox via mcp.callTool().

https://claude.ai/code/session_01GZSn58VARmT7FtYkrstnxS
@github-actions
Copy link
Copy Markdown

github-actions bot commented Mar 3, 2026

MCP-V8 Load Test Benchmark Report

Comparison of single-node vs 3-node cluster at various request rates.

Results

Topology Target Rate Actual Iter/s HTTP Req/s Exec Avg (ms) Exec p95 (ms) Exec p99 (ms) Success % Dropped Max VUs
cluster-stateful 100/s 45.8 1889 2022.27 5131.55 5154.56 70.8% 3027 100
cluster-stateful 200/s 48.6 3725.4 3858.97 5154.57 5186.16 39.9% 8838 200
cluster-stateless 1000/s 345.8 708.4 2700.41 10001.46 12936.42 61.7% 37823 1000
cluster-stateless 100/s 99.8 388.4 98.57 107.02 109.81 100% 2 12
cluster-stateless 200/s 199.5 785 102.05 109.87 115.29 100% 7 27
cluster-stateless 500/s 345.1 1391.3 1301.33 4595.85 7361.32 100% 8907 500
single-stateful 100/s 20.2 1879.9 4670.17 5119.89 5133.37 74.7% 4687 100
single-stateful 200/s 38.1 3749 4973.8 5132.6 5139.93 5.2% 9522 200
single-stateless 1000/s 122.3 240.5 7501.88 11615.11 15834.52 56.3% 51802 1000
single-stateless 100/s 99.6 416 117.87 164.5 214.8 100% 11 21
single-stateless 200/s 150.1 406.5 1107.35 2769.04 3782.51 100% 2859 200
single-stateless 500/s 105.6 284.4 4541.27 9936.39 13110.77 97.8% 23455 500

P95 Latency

Topology Rate P95 (ms)
cluster-stateful 100/s 5131.55 ███████████████████████████
cluster-stateful 200/s 5154.57 ███████████████████████████
cluster-stateless 100/s 107.02 ███████████████
cluster-stateless 200/s 109.87 ███████████████
cluster-stateless 500/s 4595.85 ███████████████████████████
cluster-stateless 1000/s 10001.46 ██████████████████████████████
single-stateful 100/s 5119.89 ███████████████████████████
single-stateful 200/s 5132.6 ███████████████████████████
single-stateless 100/s 164.5 ████████████████
single-stateless 200/s 2769.04 █████████████████████████
single-stateless 500/s 9936.39 █████████████████████████████
single-stateless 1000/s 11615.11 ██████████████████████████████

Notes

  • Target Rate: The configured constant-arrival-rate (requests/second k6 attempts)
  • Actual Iter/s: Achieved iterations per second (each iteration = 1 POST /api/exec)
  • HTTP Req/s: Total HTTP requests per second (1 per iteration)
  • Dropped: Iterations k6 couldn't schedule because VUs were exhausted (indicates server saturation)
  • Topology: single = 1 MCP-V8 node; cluster = 3 MCP-V8 nodes with Raft

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants