[mcp-analysis] MCP Structural Analysis - 2026-09-25

[github-actions[bot]]

Daily structural analysis of GitHub MCP tool responses across 10 representative tools/toolsets. Today's average usefulness rating was 3.60/5 — github_context and list_workflows (mcpscripts wrapper) tied for best (5/5), list_issues scored worst (2/5) due to an integrity-policy redaction blocking visibility into a filtered item. The code_security toolset remains the heaviest by far, averaging ~7,187 tokens/call across 18 measured days, and spiked to ~7,300 tokens today from two newly-surfaced CodeQL alerts.

Key Metrics (Today)

Metric	Value
Tools Analyzed	10
Total Tokens (Today)	10,962
Average Usefulness Rating	3.60/5
Best Rated Tool	github_context & list_workflows: 5/5
Worst Rated Tool	list_issues: 2/5

Full Structural Analysis Report

Usefulness Ratings for Agentic Work

Tool	Toolset	Rating	Assessment
github_context	workflow_context	⭐⭐⭐⭐⭐	Excellent — zero-cost injected identity block
list_workflows (mcpscripts)	actions	⭐⭐⭐⭐⭐	Excellent — clean JSON, explicit pagination, no bloat
get_file_contents	repos	⭐⭐⭐⭐	Good — complete file content, icon overhead dominates payload
list_pull_requests	pull_requests	⭐⭐⭐⭐	Good — flat, minimal, directly actionable
search_code	search	⭐⭐⭐⭐	Good — inline code fragment avoids follow-up call, but rate-limit prone
list_code_scanning_alerts	code_security	⭐⭐⭐	Adequate — usable but each alert repeats large boilerplate help text
get_label	labels	⭐⭐⭐	Adequate — tiny payload buried in fixed icon _meta overhead
list_discussions	discussions	⭐⭐⭐	Adequate — good pagination metadata, same icon overhead pattern
search_users	users	⭐⭐⭐	Adequate — minimal identity fields, needs follow-up for profile detail
list_issues	issues	⭐⭐	Limited — integrity-policy redaction hides the one real result

Schema Analysis

Tool	Type	Depth	Key Fields	Notes
github_context	markdown	1	actor, repository, workspace, workflow-run-id	Prompt-injected, no MCP call
list_workflows	object	2	total_count, per_page, workflows[].id/name/path/state	Best-in-class wrapper schema
get_file_contents	object	3	content.resource.text, content.text, _meta.serverInfo	Icon block ~75% of payload
list_pull_requests	array	1	number, title, state, html_url	Flat array, no nesting overhead
search_code	object	2	items[].name/path, text_matches[].fragment, total_count	Inline match fragments
list_code_scanning_alerts	array	3	number, rule.id/severity/full_description/help, most_recent_instance.location	No icon overhead, but verbose per-alert help text
get_label	object	1	name, color, description, id	Simplest schema, wrapped in icon overhead
list_discussions	object	3	discussions[].category.name, html_url, title, user.login, pageInfo, totalCount	Nested category/user relationships
search_users	object	2	items[].login/id/profile_url/avatar_url, total_count	No name/bio/role fields
list_issues	array	2	number, title, state, labels	Integrity-policy redaction observed every day (19/19)

Response Size Analysis (18-day average)

Toolset	Avg Tokens	Tools Tested
code_security	7,187	18
repos	2,084	18
labels	830	18
discussions	710	18
search	633	18
users	627	18
actions	171	18
issues	93	18
pull_requests	88	18
workflow_context	86	18

Tool-by-Tool Analysis (Today)

Tool	Toolset	Tokens	Schema	Rating	Notes
github_context	workflow_context	65	markdown/depth 1	⭐⭐⭐⭐⭐	Zero-cost, consistent 19/19 days
list_workflows	actions	172	object/depth 2	⭐⭐⭐⭐⭐	total_count unchanged at 375
get_file_contents	repos	655	object/depth 3	⭐⭐⭐⭐	Same CODEOWNERS file as 2026-09-24
list_pull_requests	pull_requests	45	array/depth 1	⭐⭐⭐⭐	PR #63410, spec-extractor update
search_code	search	700	object/depth 2	⭐⭐⭐⭐	429 on first attempt, succeeded on retry
list_code_scanning_alerts	code_security	7,300	array/depth 3	⭐⭐⭐	Two new CodeQL alerts (943, 941) drove size up from ~5,950 baseline
get_label	labels	610	object/depth 1	⭐⭐⭐	Unchanged "bug" label lookup
list_discussions	discussions	730	object/depth 3	⭐⭐⭐	totalCount 10,916, up from 10,886
search_users	users	620	object/depth 2	⭐⭐⭐	Unchanged pelikhan lookup
list_issues	issues	65	array/depth 2	⭐⭐	Issue #63421 redacted by integrity policy

30-Day Trend Summary

Metric	Value
Data Points	180 (18 days × 10 tools, oldest day pruned for exceeding 30-day window)
Average Daily Tokens	~12,508
Average Rating Trend	Stable (first-5-days avg 3.56 vs last-5-days avg 3.54)

Recommendations

• High-value tools (rating 4-5): github_context, list_workflows (mcpscripts wrapper), get_file_contents, list_pull_requests, search_code — prefer these first when building agentic workflows.
• Tools needing improvement: list_issues (integrity-policy redaction has blocked visibility into at least one item every single measured day — worth investigating why this repository consistently has a lower-integrity issue in the most recent results), list_code_scanning_alerts (per-alert boilerplate full_description/help duplication inflates token cost — a summary-only mode would help).
• Context-efficient tools (low tokens, high rating): github_context (65 tokens, zero MCP cost), list_workflows via mcpscripts wrapper (172 tokens, no icon bloat), list_pull_requests (45 tokens).
• Context-heavy tools: list_code_scanning_alerts (7,300 tokens today, ~7,187 avg) remains the single heaviest call by a wide margin — always paginate with perPage: 5 or lower and prefer summary fields when available.

Visualizations

Response Size by Toolset

Usefulness Ratings

Daily Token Trend

Size vs Usefulness

Generated by 🔍 GitHub MCP Structural Analysis · claude · agent · 148.2 AIC · ⌖ 9.75 AIC · ⊞ 11.7K · ◷

• expires on Sep 26, 2026, 4:02 AM UTC-08:00

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-09-26T12:02:14.475Z.

Closed by Workflow