Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
Python aicode-complexitycode-healthcode-intelligencecode-qualitydead-codedeveloper-toolsdocumentationgit-analyticsmcpopen-sourcepythonrefactoringstatic-analysistechnical-debt
import type {
ChatContext,
ChatContextKind,
ChatSuggestion,
} from "@repowise-dev/types/chat";
export type {
ChatContext,
ChatContextKind,
ChatContextTargetKind,
} from "@repowise-dev/types/chat";
export interface ChatContextPresentation {
placeholder: string;
suggestions: readonly ChatSuggestion[];
}
/** The copy tables below stay plain strings so they read as copy. This is the
* one place they become the static tier. */
interface PresentationCopy {
placeholder: string;
suggestions: readonly string[];
}
function toPresentation(copy: PresentationCopy): ChatContextPresentation {
return {
placeholder: copy.placeholder,
suggestions: copy.suggestions.map((text) => ({
text,
source: "static" as const,
})),
};
}
// Homomorphic so a partial copy table stays partial: the collection tier
// deliberately covers only the kinds that read differently without a target.
function toPresentations<T extends Record<string, PresentationCopy | undefined>>(
table: T,
): { [K in keyof T]: ChatContextPresentation } {
return Object.fromEntries(
Object.entries(table)
.filter((entry): entry is [string, PresentationCopy] => entry[1] !== undefined)
.map(([kind, copy]) => [kind, toPresentation(copy)]),
) as { [K in keyof T]: ChatContextPresentation };
}
/** The empty chat's questions: what chat can draw, for readers of either kind.
* The empty state shows the first four. */
export const EMPTY_CHAT_SUGGESTIONS: readonly string[] = [
"Chart code health by package as a treemap",
"Which files are riskiest to change, and why?",
"Show what kinds of changes we made each week",
"Give me a one-slide health summary with charts",
"Where are the biggest test coverage gaps?",
"What dead code can be safely removed?",
];
const PRESENTATION_COPY: Record<ChatContextKind, PresentationCopy> = {
repository: {
placeholder: "Ask about this repository, or paste a file path",
suggestions: [
"Give me an overview of this codebase",
"Chart code health by package as a treemap",
"Which files have the worst code health?",
"What dead code can be safely removed?",
],
},
overview: {
placeholder: "Ask about this repository overview",
suggestions: [
"Explain the main architectural boundaries",
"Which parts of this repository deserve attention first?",
"Give me a one-slide health summary with charts",
],
},
documentation: {
placeholder: "Ask about the documentation on this page",
suggestions: [
"Explain this section using the source code",
"Which source files support this documentation?",
"What important details or limitations should I know?",
],
},
architecture: {
placeholder: "Ask about this architecture view",
suggestions: [
"Explain the main component boundaries",
"Chart the architecture layers by file count",
"Trace a request through this architecture",
],
},
graph: {
placeholder: "Ask about relationships in this graph",
suggestions: [
"Explain the most important relationships shown here",
"Which nodes have the widest structural impact?",
"What is omitted or capped in this view?",
],
},
health: {
placeholder: "Ask about these code health findings",
suggestions: [
"Show health bands as a donut",
"Which finding should I address first and why?",
"Propose a safe refactoring sequence",
],
},
refactoring: {
placeholder: "Ask about this repository's refactoring targets",
suggestions: [
"Turn this into a safe refactoring plan",
"Which refactoring opportunity is safest to start with?",
"Chart open findings by severity and package",
],
},
file: {
placeholder: "Ask about this file",
suggestions: [
"Explain this file's responsibility",
"Who calls into this file and what does it depend on?",
"What is risky about changing this file?",
],
},
symbol: {
placeholder: "Ask about this symbol",
suggestions: [
"Explain what this symbol does",
"Show its callers and important dependencies",
"What behavior should tests protect here?",
],
},
module: {
placeholder: "Ask about this module",
suggestions: [
"Explain this module's responsibilities",
"Which modules depend on it?",
"Where are its highest-risk boundaries?",
],
},
commit: {
placeholder: "Ask about this commit",
suggestions: [
"Summarize the intent and impact of this commit",
"Which files in this change deserve the closest review?",
"What tests should validate this change?",
],
},
contributor: {
placeholder: "Ask about this contributor's ownership context",
suggestions: [
"Summarize this contributor's ownership areas",
"Where is knowledge concentrated around their work?",
"Which files have the lowest ownership resilience?",
],
},
decision: {
placeholder: "Ask about this architectural decision",
suggestions: [
"Explain why this decision was made",
"Show the evidence and affected code",
"Has later work superseded or conflicted with it?",
],
},
risk: {
placeholder: "Ask about the risk evidence on this page",
suggestions: [
"Top 10 files by bug fixes, rest as Other",
"Explain the highest-risk result in plain language",
"Which tests reduce the most uncertainty?",
],
},
"dead-code": {
placeholder: "Ask about this dead-code result",
suggestions: [
"Is this safe to remove, and what still reaches it?",
"Dead code by kind and confidence tier",
"What should I check before deleting it?",
],
},
"blast-radius": {
placeholder: "Ask about the reach of this change",
suggestions: [
"What breaks if I change this?",
"Which dependents deserve the closest review?",
"Which tests cover this reach?",
],
},
security: {
placeholder: "Ask how this repository tracks security evidence",
suggestions: [
"How does this repository detect and track security findings?",
"Explain the evidence without overstating certainty",
],
},
usage: {
// Usage and billing figures live on the page, not behind any chat tool.
placeholder: "Ask about this repository (usage figures are on the page)",
suggestions: [
"Give me an overview of this codebase",
"Which files have the worst code health?",
"What dead code can be safely removed?",
],
},
settings: {
placeholder: "Ask about this repository configuration",
suggestions: [
"Explain the settings on this page",
"Which settings affect indexing quality?",
"What should I verify before changing this configuration?",
],
},
chat: {
placeholder: "Ask a follow-up, or paste a file path",
suggestions: EMPTY_CHAT_SUGGESTIONS,
},
};
const COLLECTION_PRESENTATION_COPY: Partial<
Record<ChatContextKind, PresentationCopy>
> = {
documentation: {
placeholder: "Ask about this repository's documentation",
suggestions: [
"Which documentation should I read first?",
"Where should a new contributor start?",
],
},
file: {
placeholder: "Ask about files in this repository",
suggestions: [
"Which files are the main entry points?",
"Which files have the worst code health?",
],
},
symbol: {
placeholder: "Ask about symbols in this repository",
suggestions: [
"Which symbols have the widest impact?",
"Show the most important public interfaces",
],
},
module: {
placeholder: "Ask about modules in this repository",
suggestions: [
"Explain the main module boundaries",
"Which modules are most tightly coupled?",
"Where should a new feature live?",
],
},
commit: {
placeholder: "Ask about repository history",
suggestions: [
"Heatmap of changes by week and category",
"Review the risk of the latest commit",
"Find the history behind an architectural choice",
],
},
contributor: {
placeholder: "Ask about ownership and contributors",
suggestions: [
"Where is repository knowledge concentrated?",
"Which areas have the lowest ownership resilience?",
"Who knows the highest-risk files best?",
],
},
decision: {
placeholder: "Ask about architectural decisions",
suggestions: [
"Summarize the active architectural decisions",
"Which decisions affect the most code?",
"Find conflicting or superseded decisions",
],
},
};
const PRESENTATIONS = toPresentations(PRESENTATION_COPY);
const COLLECTION_PRESENTATIONS = toPresentations(COLLECTION_PRESENTATION_COPY);
export function getChatContextPresentation(
context?: ChatContext,
): ChatContextPresentation {
if (context && !context.target) {
const collectionPresentation = COLLECTION_PRESENTATIONS[context.kind];
if (collectionPresentation) return collectionPresentation;
}
return PRESENTATIONS[context?.kind ?? "repository"];
}
const LEGACY_ARTIFACT_TYPE_BY_TOOL: Readonly<Record<string, string>> = {
get_overview: "overview",
get_context: "wiki_page",
get_risk: "risk_report",
get_change_risk: "risk_report",
get_why: "decisions",
search_codebase: "search_results",
get_dead_code: "dead_code",
get_dependency_path: "graph",
get_architecture_diagram: "diagram",
};
/**
* Restores artifact affordances for stored conversations whose legacy wire
* shape persisted only the tool name and result. Unknown tools deliberately
* use the generic renderer so evidence remains inspectable.
*/
export function getLegacyChatArtifactType(toolName: string): string {
return LEGACY_ARTIFACT_TYPE_BY_TOOL[toolName] ?? "generic";
}