[mcp-analysis] MCP Structural Analysis - 2026-06-30 #42479
Closed
Replies: 1 comment
-
|
This discussion has been marked as outdated by GitHub MCP Structural Analysis. A newer discussion is available at Discussion #42750. |
Beta Was this translation helpful? Give feedback.
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.
-
GitHub MCP Structural Analysis — 2026-06-30
Tested 12 representative tools across 11 toolsets against
github/gh-aw. Average usefulness 4.00/5 today; total measured payload 7,111 tokens. Six tools earned a perfect 5/5 for agentic work (get_label,list_branches,list_discussions,search_users,search_repositories,actions_list).Notable this run: the
actionsanduserstoolsets were reachable for the first time in weeks (prior runs hit permission-denied onactions).get_mestill fails with a 403 under the integration token. The two heaviest responses —list_code_scanning_alerts(3,336 tok) andlist_issues(2,164 tok) — together account for 77% of today's total tokens, driven by embeddedrule.helpmarkdown and unbounded issue bodies respectively.get_label, 40 tok)get_me: 1/5 (403) ·list_code_scanning_alerts: 2/5 (bloat)Recommendations
get_label,list_branches,search_repositories,list_discussions,search_users,actions_list,get_file_contents,list_commits,list_pull_requests.list_code_scanning_alerts— strip/optionalrule.help&full_descriptionmarkdown (≈80% of payload, duplicated across alerts sharing a rule);list_issues— add a body-excerpt/truncation flag;get_me— unusable under the integration token, agents should read identity fromgithub-contextinstead.list_branches(30 tok),get_label(40),search_users(65),search_repositories(117),actions_list(132),list_discussions(151) — all 5/5.list_code_scanning_alerts(3,336),list_issues(2,164),get_file_contents(scales with file size).Full Structural Analysis Report
Executive Summary
get_label: 5/5 @ 40 tok (most efficient of the 5/5 group)get_me: 1/5 (403 Resource not accessible by integration)Usefulness Ratings for Agentic Work
get_labellist_branchesactions_listlist_discussionssearch_userssearch_repositoriesget_file_contentslist_commitslist_pull_requestslist_issueslist_code_scanning_alertsget_meSchema Analysis
get_meget_file_contentslist_commitslist_brancheslist_issueslist_pull_requestsactions_listmethodparamlist_code_scanning_alertslist_discussionsget_labelsearch_userssearch_repositoriesResponse Size Analysis (today)
Tool-by-Tool Analysis
30-Day Trend Summary
30-day per-toolset average response size (tokens): code_security 2,156 · issues 1,792 · repos 1,222 · pull_requests 968 · actions 432 · search 372 · discussions 180 · labels 124 · users 65 · context 23.
Visualizations
Usefulness Ratings by Toolset
Average Response Size by Toolset
Daily Token Trend (30-day window)
Size vs Usefulness
References: §28444453172
Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
awmgmcpgSee Network Configuration for more information.
Beta Was this translation helpful? Give feedback.
All reactions