pipeline-debug

To diagnose SOMA pipeline failures, run a structured multi-dimensional check across TI, HW, CR, and Score Analyzer stages. The pipeline-debug Skill maps execution evidence to root causes using explicit IF-THEN rules and auto-applies trivial fixes, generating a prioritized remediation report.

SOMA pipeline degradation or timeouts occur due to failures in the TI, HW, CR, or Score Analyzer stages. By running structured diagnostics via as_get_recent_executions and as_find_broken_flows, the root-cause mapping identifies the exact stage failure and provides evidence-based remediation steps.

The best way to find root causes for broken SOMA pipeline flows is using explicit IF-THEN rule-based mapping. This diagnostic approach replaces 5–8 manual MCP calls with a single investigation pipeline, leveraging as_find_broken_flows to classify root causes and auto-apply trivial fixes.

Yes, you can automate fixing trivial SOMA pipeline errors. When pre-conditions are met, the diagnostic pipeline auto-applies trivial fixes and maintains a log trail, accelerating remediation while still escalating complex issues to a DEEP analysis report.

Yes, pipeline-debug operates within the MCP ecosystem and requires tools like as_get_recent_executions and as_find_broken_flows. These tools provide the execution history and broken flow visibility needed to produce root-cause classifications and actionable remediation instructions for the SOMA pipeline.

You should escalate a SOMA pipeline diagnostic to DEEP analysis when the initial structured multi-dimensional checks identify complex root causes that cannot be resolved by auto-applied trivial fixes, requiring further investigation beyond the standard prioritized debug report.