Automate GitHub repositories, issues, pull requests, branches, CI/CD, and permissions via Rube MCP (Composio). Manage code workflows, review PRs, search code, and handle deployments programmatically.
89
Quality
91%
Does it follow best practices?
Impact
81%
1.32xAverage score across 3 eval scenarios
PR merge safety workflow
RUBE_SEARCH_TOOLS first
0%
0%
RUBE_MANAGE_CONNECTIONS setup
0%
100%
FIND then GET PR sequence
70%
100%
Mergeable check before merge
50%
60%
CI check before merge
70%
100%
Explicit user confirmation
100%
100%
merge_method param present
100%
100%
Draft/closed handling
100%
87%
422 error handling documented
28%
42%
Connection ACTIVE guard
0%
100%
Correct tool names used
0%
100%
Without context: $0.2888 · 1m 53s · 14 turns · 15 in / 6,429 out tokens
With context: $0.5471 · 2m 31s · 22 turns · 28 in / 7,942 out tokens
Issue pagination and PR distinction
RUBE_SEARCH_TOOLS first
0%
0%
RUBE_MANAGE_CONNECTIONS setup
0%
12%
pull_request field check
100%
100%
per_page set to 100
100%
100%
Pagination loop
100%
100%
issue_audit_results.json produced
100%
100%
Silent drop warning noted
0%
0%
repos.txt input parsed
100%
100%
Issues vs PRs documented
100%
100%
Correct list tool used
100%
100%
Without context: $0.4036 · 1m 51s · 20 turns · 27 in / 6,233 out tokens
With context: $0.4924 · 1m 47s · 22 turns · 274 in / 6,214 out tokens
Branch creation and workflow dispatch
RUBE_SEARCH_TOOLS first
0%
0%
RUBE_MANAGE_CONNECTIONS setup
0%
0%
Branch SHA lookup
100%
100%
refs/ format for ref param
100%
100%
CREATE vs UPDATE distinction
0%
100%
Workflow ID via LIST_WORKFLOWS
100%
100%
Filename-only workflow_id
40%
70%
workflow_dispatch trigger requirement
100%
100%
GITHUB_CREATE_A_WORKFLOW_DISPATCH_EVENT used
75%
100%
ref param in workflow dispatch
100%
100%
Inputs limit noted
0%
100%
Without context: $0.6151 · 2m 19s · 29 turns · 32 in / 7,341 out tokens
With context: $0.3514 · 1m 31s · 15 turns · 267 in / 5,271 out tokens
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.