repo-dive 0.7.0 → 0.8.0

package/README.md

@@ -1,14 +1,16 @@
 # repo-dive
 
-Dive into a git repository's history: per-commit snapshots, an indexed metrics catalog and an interactive dashboard.
+_Dive into a git repo's history: per-commit snapshots, an indexed metrics catalog and an interactive dashboard_ <!-- markdownlint-disable-line MD036 -- a tagline, not a heading -->
 
-> **Still 0.x.**
-> The pipeline works end to end and has been run against repositories with tens of thousands of commits, but interfaces, the catalog format and the collector roster still move between minor versions — pin the version if you script against it.
-> Renamed from `repo-insighter` in 0.4.0.
+[![npm version](https://img.shields.io/npm/v/repo-dive?logo=npm&color=3c7ef6&labelColor=333)](https://www.npmjs.com/package/repo-dive)
+[![npm downloads](https://img.shields.io/npm/dm/repo-dive?logo=npm&color=3c7ef6&labelColor=333)](https://www.npmjs.com/package/repo-dive)
+[![License](https://img.shields.io/badge/license-BSD--3--Clause-3c7ef6?logo=opensourceinitiative&logoColor=white&labelColor=333)](LICENSE.md)
+[![Effect](https://img.shields.io/badge/Effect-v4-3c7ef6?logo=effect&logoColor=white&labelColor=333)](https://effect.website)
+[![MCP](https://img.shields.io/badge/MCP-ready-3c7ef6?logo=modelcontextprotocol&logoColor=white&labelColor=333)](#ai-agents-mcp)
 
 ## What it does
 
-Point it at any git repository and get an explorable catalog of insights derived from its history:
+Point `repo-dive` at any git repository and get an explorable catalog of insights derived from its history:
 
 ```sh
 cd /path/to/your/repo
@@ -27,10 +29,10 @@ See [docs/specs](docs/specs/README.md) for the architecture and [docs/research/p
 
 ## Usage
 
-Run from inside the repository you want to analyze (or pass `--repo /path/to/repo`):
+Run from inside the repository you want to analyze (or pass `--repo /path/to/repo`).
+Node 22.13 or newer is required.
 
 ```sh
-cd /path/to/your/repo
 npx repo-dive            # the whole pipeline: scan + index + dashboard
 npx repo-dive scan       # collect snapshots into .repo-dive/
 npx repo-dive index      # roll up into the metrics cube + dashboard data
@@ -38,10 +40,10 @@ npx repo-dive dashboard  # serve the interactive dashboard
 npx repo-dive status     # show catalog coverage
 npx repo-dive collectors # list available collectors
 npx repo-dive report     # export one shareable self-contained HTML file
+npx repo-dive mcp        # serve the cube to AI agents (Model Context Protocol)
+npx repo-dive gc         # clean up the catalog interactively
+npx repo-dive ignore     # keep other tools out of the catalog
 npx repo-dive query "SELECT metric, sum(value) FROM facts GROUP BY metric"
-npx repo-dive mcp    # serve the cube to AI agents (Model Context Protocol)
-npx repo-dive gc     # clean up the catalog interactively
-npx repo-dive ignore # keep other tools out of the catalog
 ```
 
 `scan` walks the repository's history and runs collectors against every commit (or a sample, per collector), writing raw snapshots into a `.repo-dive/` catalog inside the analyzed repo.
@@ -53,7 +55,7 @@ Collectors so far:
 - **churn** — lines added/deleted per commit, by file extension
 - **file-types** — file count and bytes per extension at each commit's tree
 - **directives** — eslint-disable comments by rule (block disables tracked as gray areas) and `@ts-ignore`/`@ts-expect-error`/`@ts-nocheck`
-- **dependencies** — total resolved packages from package-manager lockfiles, per package manager (pnpm, npm and yarn — classic and berry; version-aware and monorepo-aware, extensible to more managers), plus direct/dev/optional dependencies and the number of manifests read straight from `package.json` files
+- **dependencies** — resolved package totals from lockfiles, per package manager (pnpm, npm and yarn classic and berry), plus direct/dev/optional dependencies and manifest counts read straight from `package.json` files; version-aware, monorepo-aware and extensible to more managers
 - **todo-comments** — TODO/FIXME/HACK/XXX counts
 - **languages** — lines and file count per language across a commit's source files (lockfiles, minified bundles and generated data excluded)
 - **survival** — `git blame` line survival by extension, author and age cohort (sampled monthly)
@@ -61,7 +63,8 @@ Collectors so far:
 The catalog hides itself from git, but other tools that walk the repository (prettier, markdownlint, cspell, docker builds) each read one ignore file at its root.
 `scan` warns when the catalog is missing from those; `repo-dive ignore` adds it to every one of them.
 
-`index` normalizes raw snapshots into `.repo-dive/index/metrics.sqlite` (a facts-by-categories cube, rebuildable at any time) plus `dashboard.json`, and `dashboard` serves a local React app with interactive charts: languages over time, a GitHub-style commit calendar, monthly commits with AI-assisted share, churn, lint-suppression trends, dependency counts over time, code survival by cohort and author, and more.
+`index` normalizes raw snapshots into `.repo-dive/index/metrics.sqlite` — a facts-by-categories cube, rebuildable at any time — plus `dashboard.json`.
+`dashboard` then serves a local React app with interactive charts: languages over time, a GitHub-style commit calendar, monthly commits with AI-assisted share, churn, lint-suppression trends, dependency counts over time, code survival by cohort and author, and more.
 
 ## Configuration
 
@@ -98,7 +101,8 @@ export default defineConfig({
 ```
 
 `charts.weekStartsOn` sets the first day of the week in calendar-shaped charts such as the commit calendar (`"monday"` by default, `"sunday"` also supported).
-`contributors.aliases` merges the multiple identities one person commits under (work + personal email, GitHub noreply, name variants) so attribution, the contributors table and code-survival-by-contributor count them once; a group can also carry a `displayName`, a profile `url` and a `kind` (`human`/`bot`/`ai`, otherwise auto-derived — the dashboard badges bots and AI agents and lists them apart from humans).
+`contributors.aliases` merges the multiple identities one person commits under (work + personal email, GitHub noreply, name variants) so attribution, the contributors table and code-survival-by-contributor count them once.
+A group can also carry a `displayName`, a profile `url` and a `kind` — `human`/`bot`/`ai`, otherwise auto-derived, with the dashboard badging bots and AI agents and listing them apart from humans.
 `catalog.dir` moves the catalog; point it outside the repository (e.g. `"../repo-dive-catalogs/my-repo"`) to leave the analyzed working tree untouched altogether, ignore files included.
 Apart from `catalog`, which every command needs, the config is read by `index`.
 See [docs/specs/07-config.md](docs/specs/07-config.md) for details.
@@ -148,9 +152,19 @@ pnpm lint
 pnpm fix
 ```
 
+To see how heavy the published package would be:
+
+```sh
+pnpm build && pnpm report-package-size
+```
+
+It prints the tarball and unpacked sizes with a per-file breakdown, comparing them against the previous measurement.
+CI runs the same report on every push and adds it to the job summary, comparing against the latest `main`.
+
 ## Acknowledgements
 
 Thanks to [@WillJack20](https://github.com/WillJack20) for suggesting the name **repo-dive**.
+The project was published as [repo-insighter](https://www.npmjs.com/package/repo-insighter) before 0.4.0.
 
 ## License
 

package/dist/cli.js

@@ -52465,8 +52465,8 @@ var runWith = (command, config) => {
 };
 var package_default = {
 	name: "repo-dive",
-	version: "0.7.0",
-	description: "Dive into a git repository's history: per-commit snapshots, an indexed metrics catalog and an interactive dashboard",
+	version: "0.8.0",
+	description: "Dive into a git repo's history: per-commit snapshots, an indexed metrics catalog and an interactive dashboard",
 	keywords: [
 		"git",
 		"repository",
@@ -52530,6 +52530,7 @@ var package_default = {
 		"lint:tsc": "tsc --noEmit && tsc -p dashboard --noEmit",
 		"prepack": "pnpm build",
 		"prepare": "node .husky/install.mjs",
+		"report-package-size": "node scripts/report-package-size.ts",
 		"test": "vitest run"
 	},
 	"lint-staged": { "**": ["prettier --ignore-unknown --write"] },
@@ -55159,12 +55160,6 @@ var NoCollectedCommitsError = class extends TaggedError("NoCollectedCommitsError
 	}
 };
 /**
-* Heuristic for AI coding assistants appearing as commit co-authors.
-* Automation bots (renovate, dependabot, github-actions, …) are deliberately
-* not "AI": they don't reflect assisted authorship.
-*/
-var isAiCoAuthor = (coAuthor) => !/renovate|dependabot|github-actions/i.test(coAuthor) && /claude|copilot|cursor|chatgpt|openai|gemini|aider|devin|coderabbit|codegen|sweep|windsurf/i.test(coAuthor);
-/**
 * Series labels non-human contributors fold into in survival data. The
 * dashboard matches these literal strings to color the bands with the
 * reserved kind colors, so change them in both places or not at all.
@@ -55173,10 +55168,26 @@ var kindGroupLabels = {
 	bot: "Bots",
 	ai: "AI agents"
 };
-/** "Claude Fable 5 <noreply@anthropic.com>" → "Claude Fable 5" */
-var coAuthorIdentity = (coAuthor) => {
-	const angleIndex = coAuthor.indexOf("<");
-	return (angleIndex === -1 ? coAuthor : coAuthor.slice(0, angleIndex)).trim() || coAuthor.trim();
+/**
+* Splits a `Co-authored-by:` trailer into its parts:
+* `"Claude Fable 5 <noreply@anthropic.com>"` → name + email.
+*
+* Keeping the email (rather than the display name alone) is what lets a
+* co-author resolve through the very same `resolveContributor` an author goes
+* through — so alias groups, `displayName`/`url` overrides and kind derivation
+* apply identically whether a person wrote a commit or helped with one.
+*/
+var parseIdentity = (identity) => {
+	const open = identity.indexOf("<");
+	const close = identity.lastIndexOf(">");
+	if (open === -1 || close < open) return {
+		name: identity.trim(),
+		email: ""
+	};
+	return {
+		name: identity.slice(0, open).trim(),
+		email: identity.slice(open + 1, close).trim()
+	};
 };
 var sumMetric = (commit, metric, filter) => {
 	let total = 0;
@@ -55202,13 +55213,26 @@ var sumByKey = (record, keyOf) => {
 	return merged;
 };
 var buildDashboardData = (repoRoot, commits, config) => {
-	const aiCoAuthorsOf = (commit) => [...commit.factsByCollector.values()].flat().filter((fact) => fact.metric === "commits.coAuthor" && isAiCoAuthor(fact.categories?.["coAuthor"] ?? "")).map((fact) => coAuthorIdentity(fact.categories?.["coAuthor"] ?? ""));
+	/** The raw `"Name <email>"` trailers on a commit, in the order git listed them. */
+	const coAuthorsOf = (commit) => [...commit.factsByCollector.values()].flat().filter((fact) => fact.metric === "commits.coAuthor").map((fact) => fact.categories?.["coAuthor"] ?? "");
+	/**
+	* A co-author resolved the same way an author is. An identity with no email
+	* (a bare `Co-authored-by: Some Name`) keys off its name instead, so those
+	* don't all collapse into one empty-email bucket.
+	*/
+	const resolveCoAuthor = (identity) => {
+		const { name, email } = parseIdentity(identity);
+		return {
+			name,
+			resolved: config.resolveContributor(email || name, name)
+		};
+	};
 	const commitRows = commits.map((commit) => ({
 		sha: commit.sha.slice(0, 10),
 		date: commit.date,
 		author: commit.authorEmail,
 		kind: config.resolveContributor(commit.authorEmail, commit.authorName).kind,
-		ai: aiCoAuthorsOf(commit).length > 0,
+		ai: coAuthorsOf(commit).some((identity) => resolveCoAuthor(identity).resolved.kind === "ai"),
 		added: sumMetric(commit, "churn.added"),
 		deleted: sumMetric(commit, "churn.deleted")
 	}));
@@ -55287,32 +55311,80 @@ var buildDashboardData = (repoRoot, commits, config) => {
 		};
 	});
 	const contributorMap = /* @__PURE__ */ new Map();
-	for (const [index, commit] of commits.entries()) {
-		const row = commitRows[index];
-		if (!row) continue;
-		const resolved = config.resolveContributor(commit.authorEmail, commit.authorName);
-		const key = resolved.canonicalEmail.toLowerCase();
+	/**
+	* The name as spelled on this commit (author line or trailer): a configured
+	* displayName wins, otherwise the observed spelling, tidied so a bot's
+	* `[bot]` suffix doesn't double up with its kind badge.
+	*/
+	const nameOf = (resolved, observedName) => resolved.displayName ?? (observedName ? normalizeContributorName(observedName) : resolved.label);
+	/**
+	* Fetches or creates a contributor's bucket.
+	*
+	* Humans key off their canonical email alone, so an alias group folds every
+	* spelling of one person together. Bots and AI agents key off their name too:
+	* they share vendor noreply addresses — "Claude Fable 5" and "Claude Opus
+	* 4.8" are both `<noreply@anthropic.com>` — and for them the name carries the
+	* identity worth telling apart. Giving such a group a `displayName` in the
+	* config merges them back into one.
+	*/
+	const bucketFor = (resolved, observedName) => {
+		const name = nameOf(resolved, observedName);
+		const email = resolved.canonicalEmail.toLowerCase();
+		const key = resolved.kind === "human" ? email : `${name.toLowerCase()}${email}`;
 		const bucket = contributorMap.get(key) ?? {
 			email: resolved.canonicalEmail,
-			name: resolved.label,
+			name,
 			url: resolved.url,
 			kind: resolved.kind,
 			commits: 0,
 			added: 0,
-			deleted: 0
+			deleted: 0,
+			assistedBy: {},
+			assisted: {}
 		};
-		bucket.name = resolved.displayName ?? (commit.authorName ? normalizeContributorName(commit.authorName) : bucket.name);
-		bucket.commits += 1;
-		bucket.added += row.added;
-		bucket.deleted += row.deleted;
+		bucket.name = name;
 		contributorMap.set(key, bucket);
-	}
-	const contributors = [...contributorMap.values()].toSorted((left, right) => right.commits - left.commits).slice(0, 25);
-	const aiIdentityMap = /* @__PURE__ */ new Map();
-	for (const commit of commits) for (const identity of new Set(aiCoAuthorsOf(commit))) aiIdentityMap.set(identity, (aiIdentityMap.get(identity) ?? 0) + 1);
-	const aiIdentities = [...aiIdentityMap.entries()].toSorted(([, left], [, right]) => right - left).map(([identity, commitCount]) => ({
-		identity,
-		commits: commitCount
+		return bucket;
+	};
+	for (const [index, commit] of commits.entries()) {
+		const row = commitRows[index];
+		if (!row) continue;
+		const author = config.resolveContributor(commit.authorEmail, commit.authorName);
+		const authorBucket = bucketFor(author, commit.authorName);
+		authorBucket.commits += 1;
+		authorBucket.added += row.added;
+		authorBucket.deleted += row.deleted;
+		const helpers = /* @__PURE__ */ new Set();
+		for (const identity of coAuthorsOf(commit)) {
+			const { name, resolved } = resolveCoAuthor(identity);
+			if (resolved.kind !== author.kind) helpers.add(bucketFor(resolved, name));
+		}
+		const helperKinds = /* @__PURE__ */ new Set();
+		for (const helper of helpers) {
+			helper.assisted[author.kind] = (helper.assisted[author.kind] ?? 0) + 1;
+			helperKinds.add(helper.kind);
+		}
+		for (const kind of helperKinds) authorBucket.assistedBy[kind] = (authorBucket.assistedBy[kind] ?? 0) + 1;
+	}
+	const sumCounts = (counts) => Object.values(counts).reduce((total, count) => total + count, 0);
+	/** Ranking weight: authoring and helping both count as taking part. */
+	const involvementOf = (bucket) => bucket.commits + sumCounts(bucket.assisted);
+	/** Emitted as absent rather than `{}` so the JSON stays lean. */
+	const countsOrUndefined = (counts) => Object.keys(counts).length === 0 ? void 0 : counts;
+	const perKindCap = config.maxInCharts * 2;
+	const keptPerKind = {
+		human: 0,
+		bot: 0,
+		ai: 0
+	};
+	const contributors = [...contributorMap.values()].toSorted((left, right) => involvementOf(right) - involvementOf(left)).filter((bucket) => {
+		if (keptPerKind[bucket.kind] >= perKindCap) return false;
+		keptPerKind[bucket.kind] += 1;
+		return true;
+	}).map((bucket) => ({
+		...bucket,
+		assistedBy: countsOrUndefined(bucket.assistedBy),
+		assisted: countsOrUndefined(bucket.assisted)
 	}));
 	return {
 		generatedAt: (/* @__PURE__ */ new Date()).toISOString(),
@@ -55334,8 +55406,7 @@ var buildDashboardData = (repoRoot, commits, config) => {
 		dependencies,
 		topRules,
 		survival,
-		contributors,
-		aiIdentities
+		contributors
 	};
 };
 var writeSqlite = (dbPath, commits) => {