@grove-dev/astro 0.2.20 → 0.3.1

package/templates/default/public/llms-full.txt

@@ -1,206 +1,83 @@
-# example.com — full directory
+# Grove Directory — full directory
 
-Generated: 2026-06-22T16:19:40.464Z
-Records: 8
+> Generated 2026-07-01T16:50:32.236Z from 8 records.
+> Source: https://example.com/projects · Regenerate with `grove generate`.
 
----
+Each section below mirrors one record detail page.
 
-## Coolify
+## Index
 
-Open-source, self-hostable Heroku / Netlify / Vercel alternative — deploy apps and databases from a single dashboard.
-
-- **URL**: https://example.com/coolify
-- **Repository**: https://github.com/coollabsio/coolify
-- **Stack**: SvelteKit
-- **Platforms**: linux, self-host
-- **Category**: DevOps
-- **Tags**: self-host, paas, deployment, svelte
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- Self-hosted PaaS
-- VPS deployments
-- Docker-based stacks
+- [Coolify](#coolify) — DevOps · SvelteKit · 0★ — Open-source, self-hostable Heroku / Netlify / Vercel alternative — deploy apps and databases from a single dashboard.
+- [Gitea](#gitea) — DevOps · Go · 0★ — Lightweight, self-hosted Git service — painless, single-binary, low-resource alternative to GitHub.
+- [Hoppscotch](#hoppscotch) — Web · Vue · 0★ — Open-source API development ecosystem — design, test, and debug APIs faster from a single interface.
+- [InvenTree](#inventree) — Mobile · Flutter · 0★ — Open-source inventory management system with parts, stock, and purchase orders.
+- [Label Studio](#label-studio) — AI/ML · Python · 0★ — Open-source data labeling platform — annotate images, text, audio, video, and time series for ML training.
+- [Logseq](#logseq) — Web · ClojureScript · 0★ — Privacy-first, open-source knowledge base that works on top of plain text Markdown and Org-mode files.
+- [PrivateGPT](#private-gpt) — AI/ML · Python · 0★ — Interact privately with your documents using LLMs — 100% private, no data leaves your execution environment.
+- [Wakapi](#wakapi) — Mobile · Kotlin · 0★ — Open-source time tracker for developers — coding stats, language breakdowns, and project-level metrics.
 
-**Why listed:**
-- One-click app + DB deploys
-- Active dev
-- Clear upgrade path from manual Docker
+## Projects
 
----
-
-## Gitea
+### Coolify
+Open-source, self-hostable Heroku / Netlify / Vercel alternative — deploy apps and databases from a single dashboard.
+- slug: coolify
+- category: DevOps
+- stack: SvelteKit
+- stars: 0
+- url: https://example.com/projects/coolify
 
+### Gitea
 Lightweight, self-hosted Git service — painless, single-binary, low-resource alternative to GitHub.
+- slug: gitea
+- category: DevOps
+- stack: Go
+- stars: 0
+- url: https://example.com/projects/gitea
 
-- **URL**: https://example.com/gitea
-- **Repository**: https://github.com/go-gitea/gitea
-- **Stack**: Go
-- **Platforms**: linux, self-host, docker
-- **Category**: DevOps
-- **Tags**: git, forge, self-host, go
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- Self-hosted Git hosting
-- Lightweight CI/CD integration
-- Internal team forges
-
-**Why listed:**
-- Single binary, easy install
-- Mature, stable releases
-- Active community
-
----
-
-## Hoppscotch
-
+### Hoppscotch
 Open-source API development ecosystem — design, test, and debug APIs faster from a single interface.
+- slug: hoppscotch
+- category: Web
+- stack: Vue
+- stars: 0
+- url: https://example.com/projects/hoppscotch
 
-- **URL**: https://example.com/hoppscotch
-- **Repository**: https://github.com/hoppscotch/hoppscotch
-- **Stack**: Vue
-- **Platforms**: web, desktop, self-host
-- **Category**: Web
-- **Tags**: api, testing, developer-tools, vue
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- API exploration
-- REST/GraphQL/WebSocket testing
-- Self-hosted team workflows
-
-**Why listed:**
-- Lightweight, fast UI
-- PWA + desktop builds
-- Open-source Postman/Insomnia alternative
-
----
-
-## InvenTree
-
+### InvenTree
 Open-source inventory management system with parts, stock, and purchase orders.
+- slug: inventree
+- category: Mobile
+- stack: Flutter
+- stars: 0
+- url: https://example.com/projects/inventree
 
-- **URL**: https://example.com/inventree
-- **Repository**: https://github.com/inventree/InvenTree
-- **Stack**: Flutter
-- **Platforms**: ios, android, web
-- **Category**: Mobile
-- **Tags**: inventory, stock, mobile, dart, flutter
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- Inventory tracking
-- Manufacturing
-- Hardware projects
-
-**Why listed:**
-- Production-grade mobile + web stack
-- Strong community
-- Cleanly factored Dart codebase
-
----
-
-## Label Studio
-
+### Label Studio
 Open-source data labeling platform — annotate images, text, audio, video, and time series for ML training.
+- slug: label-studio
+- category: AI/ML
+- stack: Python
+- stars: 0
+- url: https://example.com/projects/label-studio
 
-- **URL**: https://example.com/label-studio
-- **Repository**: https://github.com/HumanSignal/label-studio
-- **Stack**: Python
-- **Platforms**: web, self-host
-- **Category**: AI/ML
-- **Tags**: labeling, ml-ops, annotation, python
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- ML dataset labeling
-- Multi-modal annotation
-- Active learning pipelines
-
-**Why listed:**
-- Mature, production-grade
-- Strong ML backend
-- Active commercial + community backing
-
----
-
-## Logseq
-
+### Logseq
 Privacy-first, open-source knowledge base that works on top of plain text Markdown and Org-mode files.
+- slug: logseq
+- category: Web
+- stack: ClojureScript
+- stars: 0
+- url: https://example.com/projects/logseq
 
-- **URL**: https://example.com/logseq
-- **Repository**: https://github.com/logseq/logseq
-- **Stack**: ClojureScript
-- **Platforms**: web, desktop
-- **Category**: Web
-- **Tags**: notes, knowledge-base, productivity, clojure
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- PKM (personal knowledge management)
-- Local-first note taking
-- Graph-based outliner
-
-**Why listed:**
-- Local-first with sync
-- Active community
-- Strong plugin ecosystem
-
----
-
-## PrivateGPT
-
+### PrivateGPT
 Interact privately with your documents using LLMs — 100% private, no data leaves your execution environment.
+- slug: private-gpt
+- category: AI/ML
+- stack: Python
+- stars: 0
+- url: https://example.com/projects/private-gpt
 
-- **URL**: https://example.com/private-gpt
-- **Repository**: https://github.com/zylon-ai/private-gpt
-- **Stack**: Python
-- **Platforms**: linux, macos, self-host
-- **Category**: AI/ML
-- **Tags**: llm, rag, privacy, python
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- Private document Q&A
-- Local LLM experimentation
-- Enterprise privacy use cases
-
-**Why listed:**
-- Strict no-network mode
-- Pluggable LLM providers
-- Clean ingestion pipeline
-
----
-
-## Wakapi
-
+### Wakapi
 Open-source time tracker for developers — coding stats, language breakdowns, and project-level metrics.
-
-- **URL**: https://example.com/wakapi
-- **Repository**: https://github.com/muety/wakapi
-- **Stack**: Kotlin
-- **Platforms**: android
-- **Category**: Mobile
-- **Tags**: time-tracking, analytics, mobile, kotlin
-- **Stars**: 0
-- **Last commit**: —
-
-**Best for:**
-- Personal coding stats
-- WakaTime alternative
-- Privacy-first telemetry
-
-**Why listed:**
-- Self-host friendly
-- Companion mobile app
-- Clean MVVM architecture
-
----
+- slug: wakapi
+- category: Mobile
+- stack: Kotlin
+- stars: 0
+- url: https://example.com/projects/wakapi

package/templates/default/public/llms.txt

@@ -1,112 +1,11 @@
 # Grove Directory
 
-Grove Directory is a curated, health-aware developer directory powered by
-the Grove framework. Every entry is real, runnable, and actively
-maintained — not an SDK, not a template, not a toy demo.
+A curated, health-aware developer directory powered by Grove.
 
-The directory is built from plain YAML files in the consumer's
-repository. The build pipeline renders a static Astro site. No
-database, no JavaScript framework, no CDN with sticky sessions.
+Directory: https://example.com/projects
+Records indexed: 8
+Categories: 4
 
-## Main URLs
+## Usage
 
-- Home: https://example.com/
-- Browse: https://example.com/projects
-- About: https://example.com/about
-- Contributors: https://example.com/contributors
-- Submit a record: https://example.com/submit
-
-## Filter views (one URL = one curated lens)
-
-- Trending: https://example.com/projects?label=hot
-- Recently added: https://example.com/projects?label=new
-- Established (mature): https://example.com/projects?label=mature
-- Featured: https://example.com/projects?label=featured
-- Good to learn: https://example.com/projects?lens=good-to-learn
-- Production-like: https://example.com/projects?lens=production-like
-- Beginner-friendly: https://example.com/projects?lens=beginner-friendly
-- Contribution-ready: https://example.com/projects?lens=contribution-ready
-
-## Browse axes
-
-- By stack: https://example.com/projects (use the `stack` query param)
-- By category: https://example.com/projects (use the `category` query param)
-- By platform: https://example.com/projects (use the `platform` query param)
-- By license: https://example.com/projects (use the `license` query param)
-- By health: https://example.com/projects (use the `status` query param)
-
-## What the directory contains
-
-- Real, runnable open-source records (projects, resources, entities)
-- Stack, language, and architecture at a glance
-- Activity, health signals, and maturity tiers
-- License, stars, and contribution readiness
-- Curated notes: best for, why listed, caveats
-- Filters by stack, platform, license, and status
-
-## Data layout (for AI agents that want to ingest the catalog)
-
-- `data/records/*.yml` — human-edited source of truth (one file per record)
-- `data/taxonomy/categories.yml` — 16 categories (Productivity, Finance, …)
-- `data/taxonomy/stacks.yml` — 6 stacks (Flutter, React Native, …)
-- `data/taxonomy/platforms.yml` — 7 platforms (iOS, Android, …)
-- `data/taxonomy/distribution-channels.yml` — 11 distribution channels
-- `data/generated/records.full.json` — full record projection
-- `data/generated/records.index.json` — slim list-page projection
-- `data/generated/site-config.json` — blueprint + site metadata
-
-For per-record markdown (one section per record with description, repo,
-stack, platforms, stars, last-commit), see `/llms-full.txt`.
-
-## Blueprint system
-
-Grove supports three blueprints out of the box:
-
-- `project-directory` (default) — `kind: project` records under
-  `/projects/<slug>`. Use this for code-reference showcases.
-- `resource-hub` — `kind: resource` records (datasets, libraries,
-  tools) under `/resources/<slug>`.
-- `ecosystem-map` — `kind: entity` records (companies, foundations,
-  people) under `/entities/<slug>`.
-
-The blueprint is selected in `grove.config.ts` and propagates through
-to every page, route, and component — no per-blueprint forks needed.
-
-## How the build runs
-
-```
-pnpm run build:data    # yml → JSON (grove generate)
-pnpm run build:sitemap # emit public/sitemap.xml (grove sitemap)
-pnpm run build:llms    # emit public/llms-full.txt (committed llms.txt is preserved)
-astro build            # render the static site
-```
-
-## Contributing
-
-Open a PR against `data/records/<slug>.yml`. The build pipeline will
-fail if your record doesn't validate against the blueprint schema.
-Read `grove.config.ts` and the `data/taxonomy/` files for the
-available fields and enum values.
-
-## License
-
-MIT — the directory content, schema, and source code are all
-permissively licensed. The data is yours to fork, mirror, or
-republish; attribution to the original author is appreciated but not
-required.
-
-## Contact
-
-Issues: open a ticket on the consumer repository's issue tracker.
-Pull requests: see `CONTRIBUTING.md` for the workflow.
-
-## LLM-specific guidance
-
-- Per-record markdown: `/llms-full.txt` (generated, ~2,000 lines for
-  ~150 records).
-- Sitemap: `/sitemap.xml` (lists every record's detail URL).
-- robots.txt: explicitly allows GPTBot, ChatGPT-User, ClaudeBot,
-  Claude-Web, anthropic-ai, PerplexityBot, Google-Extended,
-  Applebot-Extended, cohere-ai, CCBot, Amazonbot.
-- Schema: each detail page emits a `SoftwareSourceCode` (or
-  `Dataset` / `Organization`) JSON-LD block for AI crawlers.
+Use /llms-full.txt for record-level details. Prefer detail pages for citations.

package/templates/default/public/sitemap.xml

@@ -2,61 +2,61 @@
 <urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
   <url>
     <loc>https://example.com/</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>daily</changefreq>
     <priority>1.0</priority>
   </url>
   <url>
     <loc>https://example.com/projects</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>daily</changefreq>
     <priority>0.9</priority>
   </url>
   <url>
     <loc>https://example.com/projects/coolify</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/gitea</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/hoppscotch</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/inventree</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/label-studio</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/logseq</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/private-gpt</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>
   <url>
     <loc>https://example.com/projects/wakapi</loc>
-    <lastmod>2026-06-22T16:19:39.853Z</lastmod>
+    <lastmod>2026-07-01T16:50:32.236Z</lastmod>
     <changefreq>weekly</changefreq>
     <priority>0.7</priority>
   </url>

package/templates/default/src/data/records.ts

@@ -38,6 +38,7 @@ import { fileURLToPath } from "node:url";
 import { dirname, resolve } from "node:path";
 import { marked } from "marked";
 import sanitizeHtml from "sanitize-html";
+import { prettySlug } from "@grove-dev/astro";
 import type {
   ProjectRecord,
   ResourceRecord,
@@ -79,12 +80,39 @@ interface SiteConfigPayload {
     labelSingular?: string;
     labelPlural?: string;
   };
+  taxonomy?: Partial<
+    Record<
+      "categories" | "stacks" | "platforms" | "distributionChannels",
+      Array<{ id: string; name: string }>
+    >
+  >;
   name?: string;
 }
 
-const fullRecordsRaw: Resource[] = (fullPayload as FullPayload).records ?? [];
-const indexRecordsRaw: IndexRecord[] = (indexPayload as IndexPayload).records ?? [];
+const fullRecordsRaw: Resource[] =
+  (fullPayload as unknown as FullPayload).records ?? [];
+const indexRecordsRaw: IndexRecord[] =
+  (indexPayload as unknown as IndexPayload).records ?? [];
 const siteConfigRaw: SiteConfigPayload = siteConfigPayload as SiteConfigPayload;
+const taxonomyMaps = Object.fromEntries(
+  Object.entries(siteConfigRaw.taxonomy ?? {}).map(([kind, entries]) => [
+    kind,
+    new Map((entries ?? []).map((entry) => [entry.id, entry.name])),
+  ]),
+) as Partial<
+  Record<
+    "categories" | "stacks" | "platforms" | "distributionChannels",
+    Map<string, string>
+  >
+>;
+export const taxonomy = siteConfigRaw.taxonomy ?? {};
+
+export function taxonomyLabel(
+  kind: "categories" | "stacks" | "platforms" | "distributionChannels",
+  id: string,
+): string {
+  return taxonomyMaps[kind]?.get(id) ?? prettySlug(id);
+}
 
 /** Full records (every record, all visibility). Use this for the
  *  detail page (where you need `content`, `bestFor`, `whyListed`,
@@ -224,6 +252,10 @@ export const fullItems: Resource[] = (() => {
   return fullRecords.filter((r) => r.kind === blueprintKind);
 })();
 
+export const fullProjects: ProjectRecord[] = fullRecords.filter(
+  (record): record is ProjectRecord => record.kind === "project",
+);
+
 // ──────────────────────────────────────────────────────────────────────
 // Markdown content rendering (sanitized)
 // ──────────────────────────────────────────────────────────────────────
@@ -316,3 +348,49 @@ for (const r of fullRecords) {
 export function getContentHtml(slug: string): string | null {
   return contentHtmlBySlug.get(slug) ?? null;
 }
+
+/**
+ * Sanitized Markdown for a consumer-authored page under
+ * `content/pages/<page>.md`. Default template pages use this to
+ * accept custom copy while keeping Grove's layout and components.
+ */
+export function getPageContentHtml(page: string): string | null {
+  const candidates = [
+    resolve(here, "..", "..", "content", "pages", `${page}.md`),
+    resolve(here, "..", "..", "..", "content", "pages", `${page}.md`),
+    resolve(process.cwd(), "content", "pages", `${page}.md`),
+  ];
+  const path = candidates.find((candidate) => existsSync(candidate));
+  if (!path) return null;
+
+  try {
+    const rawHtml = marked.parse(readFileSync(path, "utf8"), { async: false }) as string;
+    return sanitizeHtml(rawHtml, {
+      allowedTags: [
+        "h1", "h2", "h3", "h4",
+        "p", "br", "hr",
+        "ul", "ol", "li",
+        "strong", "em", "b", "i", "u", "s", "del",
+        "a", "code", "pre", "blockquote", "img",
+      ],
+      allowedAttributes: {
+        a: ["href", "title", "rel", "target"],
+        img: ["src", "alt", "title", "width", "height", "loading"],
+      },
+      allowedSchemes: ["http", "https", "mailto", "tel"],
+      allowedSchemesByTag: {
+        a: ["http", "https", "mailto", "tel"],
+        img: ["http", "https"],
+      },
+      transformTags: {
+        a: sanitizeHtml.simpleTransform("a", {
+          rel: "noopener noreferrer",
+          target: "_blank",
+        }, true),
+      },
+      disallowedTagsMode: "discard",
+    });
+  } catch {
+    return null;
+  }
+}

package/templates/default/src/pages/[slug]/[recordSlug].astro

@@ -64,7 +64,7 @@ if (!record) {
 }
 
 const isProject = record.kind === "project";
-const proj: ProjectRecord | null = project;
+const proj: ProjectRecord | null = project ?? null;
 
 const name =
   record.kind === "resource" ? (record as { title: string }).title : (record as { name: string }).name;
@@ -220,18 +220,20 @@ if (isProject && proj) {
     keywords: (record.tags ?? []).length > 0 ? (record.tags ?? []).join(", ") : undefined,
     isAccessibleForFree: true,
   };
-} else {
+} else if (entity) {
   jsonLd = {
     "@context": "https://schema.org",
-    "@type": record.type === "person" ? "Person" : "Organization",
+    "@type": entity.type === "person" ? "Person" : "Organization",
     name,
     headline: name,
-    description: record.description || undefined,
+    description: entity.description || undefined,
     url: homepageUrl || undefined,
-    foundingDate: record.founded || undefined,
-    location: record.location || undefined,
-    keywords: (record.tags ?? []).length > 0 ? (record.tags ?? []).join(", ") : undefined,
+    foundingDate: entity.founded || undefined,
+    location: entity.location || undefined,
+    keywords: (entity.tags ?? []).length > 0 ? (entity.tags ?? []).join(", ") : undefined,
   };
+} else {
+  jsonLd = {};
 }
 ---