メニュー
Use when navigating or auditing the docs/ knowledge graph — explain a doc's links, find the path between two docs, or surface god nodes / bridges / orphans.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Author a culture's armor items and swap troop equipment rosters via the generator + validation pipeline. Use when adding/revamping armor. Enforces the canonical-folder + cover-attribute rules.
Give DOTS lords lore-driven skill values and traits via the DOTS SkillSet system. Use when a canonical lord has wrong stats or a culture roster needs a balance pass.
Author or revamp a DOTS culture's armor set, troop tree, and recruitment wiring end-to-end. Use for new cultures or troop-tree revamps. Follows docs/ai-includes/new-culture-authoring.md.
Use when reviewing an external repo or article to adopt practices into DOTS — security-vet first, map novel vs duplicative, port (never install), review, commit your changes.
Structured 4-phase self-debug for failing agent runs (looping, drifting, burning tokens). Capture, diagnose, contained recovery, report. Complements /investigate (which is for code bugs); this skill is for harness/agent failures.
Incrementally fix dotnet build errors with minimal diffs, one error at a time
| name | doc-graph |
| description | Use when navigating or auditing the docs/ knowledge graph — explain a doc's links, find the path between two docs, or surface god nodes / bridges / orphans. |
| allowed-tools | Bash, Read |
| argument-hint | ["explain <doc> | path <a> <b> | metrics"] |
Query + audit DOTS's documentation knowledge graph via tools/graph_query.py. This is ADR-010 Phase 5 (graph analytics over the link graph). Authoritative reference + interpretation guide: docs/features/doc-graph.md — read it for what the metrics mean and the scope boundaries; this skill is just the entry point.
The tool is deterministic and offline (reads only the markdown links already in docs/). .md nodes only — Serena owns C# symbols, the dots-moduledata MCP owns game-data refs.
Two modes — Navigate (mid-task, token-cheap) and Audit (periodic KB hygiene). Concrete triggers:
explain Xpath A B --directedexplain X/knowledge-compile, feature wave) — did any ship orphaned? → metrics/skill-stocktake for docs → metrics --top 15Not for full-text search (grep / INDEX.md), C# symbols (Serena), or game-data refs (dots-moduledata MCP). Full use-case table + "when NOT to use it": docs/features/doc-graph.md.
python tools/graph_query.py explain <doc> # inbound + outbound neighbours
python tools/graph_query.py path <a> <b> --directed # forward link chain (prefer --directed)
Name a doc by slug (career-system), filename, or repo-relative path. Use --directed for path — undirected routes through the INDEX.md super-hub and reports a misleading "2 hops through INDEX" for almost any pair.
python tools/graph_query.py metrics --top 15 (or --summary for counts only). Read the output.INDEX.md — features/X.md means X has no peer cross-links; reinforce it from a sibling doc/RCA..md link; link into INDEX/a sibling, or delete.tools/build_backlinks.py + /lint-docs to refresh + confirm, and re-run metrics to verify the signal moved.--infer / memory-layer / MCP ideas — they're documented as future phases in the feature doc and were intentionally not built.python tools/graph_query.py … directly (verbs accept --json).