Debug Issue

by tirth82056b12d11625cbec3b6773e076cb3d136464fa90e5No licenseListed Oct 9, 2026Updated Oct 9, 2026

Systematically debug issues using graph-powered code navigation

AI-generated overview

Traces bugs through a code knowledge graph before reading source, using graph query tools to narrow scope.

What it does: This skill guides an agent through a systematic debugging workflow that uses graph-powered code navigation tools. It directs the agent to run semantic search for issue-related code, trace callers and callees, follow execution flows to the trigger point, check recent changes, and assess the impact radius of a fix. It also sets token-efficiency rules, including a minimal-context call first, minimal detail levels, and a budget of about five tool calls and 800 tokens of graph output per task. It produces a narrowed scope for the bug, but explicitly notes the graph does not replace reading the implementation and its tests.
When to use it: Use it when debugging a reported issue in a codebase that is indexed in a knowledge graph and the relevant graph tools are available. It fits situations where you want to locate the suspect code, trace how execution reaches it, and understand what a fix might affect before editing source.
Requirements: Requires access to graph tools such as semantic_search_nodes_tool, query_graph_tool, get_flow_tool, detect_changes_tool, get_impact_radius_tool, and get_minimal_context_tool, plus a codebase indexed in that graph. It ships no scripts; it is instructions only.

Trace a bug through the knowledge graph before reading source.

Call semantic_search_nodes_tool to find code related to the issue.
Call query_graph_tool with callers_of and callees_of to trace the call chain in both directions.
Call get_flow_tool for the execution path that reaches the suspect code. Its entry point is where the bug is triggered.
Call detect_changes_tool to check whether a recent change caused the issue.
Call get_impact_radius_tool on the suspect files to see what a fix would affect.

Call get_minimal_context_tool(task="<your task>") before any other graph tool.
Pass detail_level="minimal" wherever a tool accepts it. Use "standard" only when minimal is not enough.
Prefer a targeted query_graph_tool call over a broad listing call.
Budget: about five tool calls and 800 tokens of graph output per task.
Read the implementation and its tests before changing code. The graph narrows scope; it does not replace the source.

Source and attribution

License: No license

Content belongs to its original authors. SourceWeft indexes it from a public repository.