메뉴
Use when you need to ask questions about a codebase or understand code using a knowledge graph
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | understand-chat |
| description | Use when you need to ask questions about a codebase or understand code using a knowledge graph |
| argument-hint | ["query"] |
Answer questions about this codebase using the knowledge graph at .understand-anything/knowledge-graph.json.
The knowledge graph JSON has this structure:
project — {name, description, languages, frameworks, analyzedAt, gitCommitHash}nodes[] — each has {id, type, name, filePath?, summary, tags[], complexity, languageNotes?}
file:path, function:path:name, config:path, article:pathedges[] — each has {source, target, type, direction, weight}
layers[] — each has {id, name, description, nodeIds[]}tour[] — each has {order, title, description, nodeIds[]}Check that .understand-anything/knowledge-graph.json exists in the current project root. If not, tell the user to run /understand first.
Read project metadata only — use Grep or Read with a line limit to extract just the "project" section from the top of the file for context (name, description, languages, frameworks).
Search for relevant nodes — use Grep to search the knowledge graph file for the user's query keywords: "$ARGUMENTS"
"name" fields: grep -i "query_keyword" in the graph file"summary" fields for semantic matches"tags" arrays for topic matchesid values of all matching nodesFind connected edges — for each matched node ID, Grep for that ID in the edges section to find:
Read layer context — Grep for "layers" to understand which architectural layers the matched nodes belong to.
Answer the query using only the relevant subgraph:
Throwaway HTML mockups: 2-3 design variants to compare.
Breaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible.
Execute implementation plans by dispatching fresh subagents per task with two-stage review. Use when executing a multi-task implementation plan, when context is getting too long for quality output, or when you want maximum isolation between tasks.
Core tGD rules that MUST be followed at all times — the Verification Iron Law, per-phase tone, the Command Closing Report, and human sign-off. Load this at the start of every tGD session (the session-start meta-skill tgd-router points here) and whenever a command references it. Do not skip, do not rationalize exceptions.
Discovers and invokes agent skills. Use when starting a session or when you need to discover which skill applies to the current task. This is the meta-skill that governs how all other skills are discovered and invoked.
Creates specs before coding. Use when starting a new project, feature, or significant change and no specification exists yet. Use when requirements are unclear, ambiguous, or only exist as a vague idea.