kiro-debug

Diagnose implementation failures through root-cause-first investigation and structured debug reports.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/scapia-oss/compass --skill kiro-debug-scapia-oss
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kiro-debug
Source: https://github.com/scapia-oss/compass/tree/main/tools/cc-sdd/templates/agents/claude-code-skills/skills/kiro-debug
Command: npx skills add https://github.com/scapia-oss/compass --skill kiro-debug-scapia-oss

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When implementation tasks fail, verification breaks, or repeated fixes do not converge, teams often fall into guess-first patching that hides the real cause. This Skill enforces evidence-driven root cause analysis before any fix is proposed, preventing rework and premature escalation. ## Core Features & Use Cases - Root Cause Classification: Categorizes failures into ten classes such as MISSING_DEPENDENCY, RUNTIME_MISMATCH, CONFIG_GAP, and SPEC_CONFLICT. - Evidence Gathering Workflow: Reads error output, inspects repository state (package.json, tsconfig, git diff), and searches official documentation and issue trackers when web access is available. - Structured Debug Report: Produces ROOT_CAUSE, FIX_PLAN, VERIFICATION commands, and a NEXT_ACTION decision (RETRY_TASK, BLOCK_TASK, or STOP_FOR_HUMAN) with a confidence level. - Learning Capture: Records generalizable traps into spec learnings and cross-spec pattern files so future specs avoid the same dead ends. - Use Case: An implementer is blocked after two failed remediation attempts on a module import error; this Skill investigates, classifies it as MODULE_FORMAT, and returns a minimal repo-fixable plan. ## Quick Start Invoke the kiro-debug skill with a failure summary such as the exact error message and failing command to receive a root cause analysis and fix plan.

Frequently Asked Questions about kiro-debug

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug a failed implementation task with an AI agent?

Provide the exact error message, failing command output, current git diff, and the task brief to the debugging skill. It reads the error, inspects repository configuration, optionally searches official docs, and returns a root cause with a minimal fix plan.

What is root-cause-first debugging?

Root-cause-first debugging gathers evidence from error output, repository state, and documentation before proposing any fix. It avoids multi-fix shotgun plans by identifying a single root cause and the smallest safe corrective action.

When should a blocked task be escalated to a human instead of retried?

Escalate when the blocker requires a product decision, external credentials, unavailable hardware, or re-scoping due to a spec conflict. If the issue is fixable inside the current approved task plan, the task should be retried instead.

Why do repeated fixes fail to converge on the same error?

Repeated failures usually mean the true cause is a task ordering problem, wrong task decomposition, or a spec conflict rather than a code bug. Classifying the failure category reveals whether the task plan itself must be revised.

Can debugging results be reused across future specs?

Yes. Generalizable traps such as environment gotchas or wrong assumptions are appended to spec learnings and promoted into cross-spec pattern files, so future investigations can check known failure modes first.