Repository navigation
[mcp-analysis] MCP Structural Analysis - 2026-10-14 #68689
Closed
Replies: 1 comment
|
This discussion was automatically closed because it expired on 2026-10-15T12:01:47.937Z.
|
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
MCP Structural Analysis — 2026-10-14
Tested 11 representative tool calls across 10 GitHub MCP toolsets (one via the
mcpscriptswrapper). Best-rated:github_context,list_pull_requests(with explicitfields), and thelist_workflowswrapper — all 5/5. Worst-rated:search_code, which hit a secondary rate limit (429) for the second time in 3 days. A recurring structural risk emerged across all three days of data: three separate toolsets (issues,code_security,search) returned empty arrays today purely because their sole matching item was removed by integrity/secrecy policy filtering — an agent cannot distinguish "nothing exists" from "something exists but was filtered."github_context,list_pull_requests,list_workflows— 5/5search_code— 1/5 (rate-limited)Full Structural Analysis Report
Usefulness Ratings for Agentic Work
github_contextlist_pull_requestsfieldsfilter yields compact, fully actionable recordslist_workflows(mcpscripts)get_file_contentslist_discussionsget_labellist_issueslist_code_scanning_alertssearch_usersget_usertool; worst payload-to-overhead ratiosearch_issuessearch_code; also filtered to emptysearch_codeSchema Analysis
github_contextget_file_contentslist_issueslist_pull_requestsfieldslist_code_scanning_alertslist_discussionsget_labelsearch_userssearch_issuessearch_codelist_workflows(mcpscripts)Response Size Analysis
Tool-by-Tool Analysis (2026-10-14)
github_context#falseplaceholdersget_file_contentslist_issueslist_pull_requestsfieldsfilter eliminated all overheadlist_code_scanning_alertslist_discussionsget_labelsearch_usersget_user; worst overhead ratiosearch_issuessearch_codefields)list_workflows30-Day Trend Summary
Visualizations
Response Size by Toolset
Usefulness Ratings
Daily Token Trend
Size vs Usefulness
Recommendations
github_context(free),list_pull_requests/other list tools with explicitfields, and themcpscripts list_workflowswrapper — prefer these patterns whenever available.list_issuesandlist_code_scanning_alertsshould surface filtered-item counts more prominently (currently a buried text notice) so agents don't silently treat policy-filtered results as "nothing found."search_codeneeds backoff/retry handling for its frequent 429s.github_context(85 tok),list_workflowswrapper (170 tok),list_pull_requestswithfields(115 tok).get_file_contents(1,380 tok for a 1KB file),list_discussions(1,010 tok),search_users/get_label(~900 tok) — all driven by a fixed ~3.4KB server icon/_meta payload attached to every stock GitHub MCP response, not by the actual data.References:
All reactions