@adminide-stack/yantra-help-browser 12.0.16-alpha.6 → 12.0.16

package/lib/templates/content/product-features/real-time-search.md

@@ -0,0 +1,84 @@
+# Real-Time Search
+
+Yantra's real-time search delivers up-to-the-minute information by crawling the web as you ask, combining fresh data with AI-powered analysis.
+
+---
+
+## How Real-Time Search Works
+
+Unlike traditional search engines that rely on pre-built indexes, Yantra can fetch and process live web content at query time:
+
+1. **Query analysis** — Yantra determines whether your question needs real-time data (e.g., "latest," "today," "current price").
+2. **Targeted crawling** — Relevant sources are fetched in real-time based on relevance and authority.
+3. **Content extraction** — Key information is extracted from the fetched pages, filtering out ads, navigation, and boilerplate.
+4. **AI synthesis** — The extracted content is analyzed and synthesized into a coherent, cited response.
+5. **Source verification** — All claims are cross-referenced across multiple sources for accuracy.
+
+> **Speed:** Real-time searches typically return results in 3–8 seconds, depending on the number of sources fetched.
+
+---
+
+## When Real-Time Search Activates
+
+Yantra automatically uses real-time search when your query involves:
+
+| Trigger             | Example                                       |
+| ------------------- | --------------------------------------------- |
+| Current events      | "What happened at the tech conference today?" |
+| Time-sensitive data | "Current stock price of NVIDIA"               |
+| Recent developments | "Latest release notes for React 19"           |
+| Live information    | "Weather in San Francisco right now"          |
+| Breaking news       | "Election results 2024"                       |
+
+For timeless or conceptual questions, Yantra uses its indexed knowledge base for faster responses.
+
+---
+
+## Source Selection
+
+### How sources are chosen
+
+Yantra prioritizes sources based on:
+
+- **Authority** — Established, reputable sources are ranked higher.
+- **Freshness** — More recently updated content is preferred for time-sensitive queries.
+- **Relevance** — Sources that closely match your specific query terms and intent.
+- **Diversity** — Multiple perspectives are gathered to avoid single-source bias.
+
+### Number of sources
+
+| Search mode   | Sources fetched |
+| ------------- | --------------- |
+| Regular       | 5–10            |
+| Pro Search    | 10–20           |
+| Deep Research | 20–50+          |
+
+---
+
+## Inline Citations
+
+Every real-time search response includes numbered inline citations:
+
+- Click a citation number `[1]` to jump to the source.
+- Hover to see a preview of the source title, domain, and publish date.
+- All sources are listed at the bottom of the response with full URLs.
+
+---
+
+## Limitations
+
+- **Paywalled content** — Yantra cannot access content behind paywalls or login walls.
+- **Dynamic web apps** — Single-page applications that require JavaScript rendering may not be fully accessible.
+- **Rate limits** — Extremely high-volume real-time queries may be throttled to ensure service stability.
+- **Accuracy** — Real-time content hasn't been fact-checked by editors. Always verify critical information from primary sources.
+
+---
+
+## Combining Real-Time with AI
+
+Real-time search becomes especially powerful when combined with Yantra's AI analysis:
+
+- **Summarize breaking news** — Get a concise overview instead of reading 10 articles.
+- **Compare sources** — Ask Yantra to compare what different outlets are reporting.
+- **Extract data** — Pull specific numbers, dates, or quotes from real-time sources.
+- **Track over time** — Use saved searches with real-time mode to monitor topics continuously.

package/lib/templates/content/product-features/saved-searches.md

@@ -0,0 +1,75 @@
+# Saved Searches
+
+Save any search query for quick access later, organize results into collections, and set up alerts for ongoing monitoring.
+
+---
+
+## Saving a Search
+
+### Quick save
+
+Click the **bookmark icon** on any search result page to save the current query. The search is added to your default collection.
+
+### Save with options
+
+1. Click the **⋯** menu on the search results page.
+2. Select **Save Search**.
+3. Choose a collection or create a new one.
+4. Optionally add tags for easy filtering.
+5. Optionally enable **monitoring** to receive alerts when new matching content appears.
+
+---
+
+## Collections
+
+Organize your saved searches into themed collections:
+
+| Feature            | Details                                                                       |
+| ------------------ | ----------------------------------------------------------------------------- |
+| Create collections | Group related searches (e.g., "Competitor Research", "Tech Stack Evaluation") |
+| Share collections  | Share with workspace members (view or edit access)                            |
+| Reorder items      | Drag and drop to prioritize                                                   |
+| Bulk actions       | Export, delete, or move multiple items at once                                |
+| Nested collections | Create sub-collections for deeper organization                                |
+
+### Managing collections
+
+- **Create** — Click **+ New Collection** in the Library sidebar.
+- **Rename** — Right-click the collection name > **Rename**.
+- **Delete** — Right-click > **Delete**. You'll be asked to confirm. Saved searches inside are moved to Uncategorized, not deleted.
+
+---
+
+## Search Monitoring
+
+Turn any saved search into a continuous monitor:
+
+1. Save a search and enable **Monitoring**.
+2. Set the check frequency: hourly, daily, or weekly.
+3. Choose notification method: in-app, email, or Slack.
+4. Yantra will re-run the search on schedule and notify you when new results appear.
+
+### Use cases for monitoring
+
+- **Company news** — Save a search for your company name and competitor names to track mentions.
+- **Technology updates** — Monitor a framework or library for new releases and CVEs.
+- **Industry trends** — Track emerging topics in your field with broad keyword searches.
+- **Job market** — Monitor specific roles or skill requirements.
+
+---
+
+## Accessing Saved Searches
+
+- **Library** — All saved searches and collections are accessible from the Library page in the sidebar.
+- **Quick access** — Press ⌘ + K (Mac) or Ctrl + K (Windows) and type the collection name.
+- **Search bar** — Your recent and saved searches appear as suggestions when you focus the search bar.
+
+---
+
+## Export and Sharing
+
+Share or export saved searches:
+
+- **Share link** — Generate a shareable link for specific searches or entire collections.
+- **Export** — Download collections in Markdown, JSON, or CSV format.
+- **Slack** — Post a saved search result to a Slack channel directly from the Library.

package/lib/templates/content/product-features/search-features.md

@@ -0,0 +1,82 @@
+# Search Features
+
+A comprehensive overview of everything Yantra's search can do — from natural language queries to advanced filters and multi-modal results.
+
+---
+
+## Natural Language Search
+
+Ask questions the way you'd ask a colleague:
+
+| Instead of                    | Try                                                           |
+| ----------------------------- | ------------------------------------------------------------- |
+| `kubernetes pod restart loop` | "Why does my Kubernetes pod keep restarting?"                 |
+| `react useEffect cleanup`     | "How do I properly clean up side effects in React useEffect?" |
+| `AWS S3 pricing calculator`   | "How much would it cost to store 500 GB on S3 for a year?"    |
+
+Yantra understands intent, context, and nuance — no need to reduce your question to keywords.
+
+---
+
+## Search Modes
+
+Yantra offers multiple search modes to match different needs:
+
+| Mode              | Speed              | Depth                | Best for                              |
+| ----------------- | ------------------ | -------------------- | ------------------------------------- |
+| **Regular**       | Instant (< 3s)     | Quick answers        | Factual lookups, simple questions     |
+| **Pro Search**    | Fast (5–10s)       | Detailed analysis    | Multi-step questions, comparisons     |
+| **Deep Research** | Thorough (30–120s) | Comprehensive report | Academic research, strategic analysis |
+
+Switch modes using the toggle below the search bar or set a default in **Settings > Preferences**.
+
+---
+
+## Smart Filters
+
+Narrow your results with powerful filters:
+
+| Filter           | Options                                   | Example                       |
+| ---------------- | ----------------------------------------- | ----------------------------- |
+| **Time range**   | Past hour, day, week, month, year, custom | "news from the past 24 hours" |
+| **Source type**  | Web, academic, code, news, social         | Academic papers only          |
+| **Domain**       | Specific websites or exclude domains      | Only results from github.com  |
+| **Language**     | 50+ languages                             | Results in French only        |
+| **Content type** | Articles, PDFs, videos, code repos        | PDFs about machine learning   |
+
+Filters can be applied via the filter bar or included naturally in your query: "Find Python tutorials from the last month on YouTube."
+
+---
+
+## Response Formatting
+
+Yantra automatically formats responses for readability:
+
+- **Prose** — Flowing paragraphs for explanatory answers.
+- **Lists** — Numbered steps for how-to answers, bullet points for enumerations.
+- **Tables** — Comparison data and multi-attribute answers rendered as clean tables.
+- **Code blocks** — Syntax-highlighted code with language detection.
+- **Math** — LaTeX equations rendered inline.
+
+You can explicitly request a format: "Give me this as a table" or "Show me the code."
+
+---
+
+## Source Citations
+
+Every response includes verifiable source citations:
+
+- **Inline citations** — Numbered references `[1]`, `[2]`, etc., linked to specific claims.
+- **Source cards** — Expandable cards at the bottom showing title, domain, date, and relevance score.
+- **One-click verify** — Click any citation to open the original source in a new tab.
+
+---
+
+## Search History
+
+All your searches are automatically saved:
+
+- **Recent searches** — Appear as suggestions when you focus the search bar.
+- **Full history** — Accessible from the sidebar under **History**.
+- **Search within history** — Search your past queries to find previous results.
+- **Privacy controls** — Clear history at any time from **Settings > Privacy**.

package/lib/templates/content/product-features/search-history.md

@@ -0,0 +1,86 @@
+# Search History
+
+Yantra automatically saves your search history so you can revisit past queries, re-run searches, and track your research over time.
+
+---
+
+## Accessing Your History
+
+### Quick access
+
+- **Recent searches** automatically appear when you click the search bar.
+- Press ⌘ + K (Mac) or Ctrl + K (Windows) to open the command palette, then type a past query.
+
+### Full history
+
+1. Click **History** in the sidebar navigation.
+2. Browse your complete search history, organized by date.
+3. Click any past search to re-run it or view the original results.
+
+---
+
+## History Management
+
+### Searching your history
+
+Use the search box at the top of the History page to find specific past queries:
+
+- Search by **keywords** — Find queries containing specific terms.
+- Search by **date** — Filter to a specific day, week, or month.
+- Search by **mode** — Filter by search mode (Regular, Pro Search, Deep Research).
+
+### Organizing history
+
+| Action             | How                      | Description                              |
+| ------------------ | ------------------------ | ---------------------------------------- |
+| Pin                | Click star icon          | Pins a search to the top of your history |
+| Delete             | Click × icon             | Removes a single entry                   |
+| Save to collection | Click bookmark icon      | Moves to a saved collection              |
+| Bulk delete        | Select multiple > Delete | Remove multiple entries at once          |
+
+---
+
+## Privacy Controls
+
+### Automatic cleanup
+
+Set search history to auto-delete after a specified period:
+
+- **Settings > Privacy > Data Retention > Search History**
+- Options: 30 days, 60 days, 90 days (default), 180 days, or Keep forever
+
+### Pause history
+
+Temporarily stop saving searches:
+
+- **Settings > Privacy > Search History** → Toggle off
+- Searches performed while history is paused are not recorded.
+
+### Clear all history
+
+Go to **Settings > Privacy > Clear Search History** to permanently delete all past queries. This action is irreversible.
+
+---
+
+## Incognito Search
+
+For searches you don't want saved:
+
+1. Press ⌘ + Shift + N (Mac) or Ctrl + Shift + N (Windows) to open an incognito search.
+2. Queries in incognito mode are not saved to your history.
+3. AI responses are not personalized based on your past activity.
+4. The session is discarded when you close the incognito tab.
+
+> **Note:** Incognito mode does not hide your activity from your organization's admin on Enterprise plans with audit logging enabled.
+
+---
+
+## Syncing Across Devices
+
+Your search history syncs automatically when you're signed in on multiple devices:
+
+- **Web** — Full history available.
+- **Desktop app** — Full history available.
+- **Mobile app** — Last 100 searches available offline; full history requires an internet connection.
+
+Sync happens in real-time. A search performed on your phone appears in your desktop history within seconds.

package/lib/templates/content/product-features/source-citations.md

@@ -0,0 +1,100 @@
+# Source Citations
+
+Every Yantra response is backed by verifiable sources. Inline citations let you trace any claim back to its origin, giving you confidence in the information you receive.
+
+---
+
+## How Citations Work
+
+### Automatic citation
+
+Yantra automatically identifies claims in its responses that are supported by specific sources and adds numbered inline citations `[1]`, `[2]`, `[3]`, etc.
+
+```
+The global AI market is projected to reach $1.8 trillion by 2030 [1],
+driven primarily by enterprise adoption of generative AI tools [2].
+```
+
+### Citation sources
+
+| Source type         | Icon | Example                                |
+| ------------------- | ---- | -------------------------------------- |
+| Web article         | 🌐   | News sites, blogs, corporate pages     |
+| Academic paper      | 📄   | arXiv, PubMed, Google Scholar          |
+| Documentation       | 📘   | Official docs, API references, guides  |
+| Code repository     | 💻   | GitHub, GitLab                         |
+| Government/official | 🏛️   | Government reports, regulatory filings |
+
+---
+
+## Viewing Citations
+
+### Inline preview
+
+Hover over any citation number to see a popup with:
+
+- Source title
+- Domain name
+- Publication date
+- Relevance score (how closely the source matches the claim)
+
+### Full source list
+
+Scroll to the bottom of any response to see the complete source list with:
+
+- Full title and URL
+- Domain authority indicator
+- Date published or last updated
+- Snippet showing the specific passage that supports the claim
+
+### One-click verification
+
+Click any citation number or source card to open the original page in a new tab.
+
+---
+
+## Citation Quality
+
+### How sources are selected
+
+| Signal            | Description                                                                |
+| ----------------- | -------------------------------------------------------------------------- |
+| Domain authority  | Priority given to established, reputable sources                           |
+| Content freshness | More recent content is preferred for time-sensitive topics                 |
+| Relevance match   | Sources must specifically support the claim, not just be topically related |
+| Cross-reference   | Claims supported by multiple sources are weighted higher                   |
+| Diversity         | Multiple perspectives are included when available                          |
+
+### Accuracy indicators
+
+- **High confidence** — Claim is supported by 3+ independent sources.
+- **Medium confidence** — Claim is supported by 1–2 sources.
+- **Yantra's analysis** — Clearly labeled when content is AI-generated inference rather than directly sourced.
+
+---
+
+## Citation Settings
+
+Customize how citations appear:
+
+| Setting            | Options                      | Default           |
+| ------------------ | ---------------------------- | ----------------- |
+| Citation style     | Inline, footnote, or endnote | Inline            |
+| Show previews      | On hover, on click, or never | On hover          |
+| Include in exports | Always, optional, or never   | Always            |
+| Academic format    | APA, MLA, Chicago, IEEE      | None (links only) |
+
+Adjust at **Settings > Preferences > Citations**.
+
+---
+
+## For Academic Use
+
+Yantra can format citations in standard academic styles:
+
+1. Click the **Cite** button on any source card.
+2. Select your citation format (APA, MLA, Chicago, IEEE, Harvard).
+3. Copy the formatted citation to your clipboard.
+4. Paste directly into your paper or bibliography manager.
+
+> **Tip:** Use Deep Research mode for academic work — it fetches more sources and provides higher-quality citations.

package/lib/templates/content/scope-api/api-overview.md

@@ -0,0 +1,104 @@
+# Scope API Overview
+
+The Scope API lets you integrate Yantra's search and AI capabilities directly into your own applications, workflows, and tools.
+
+---
+
+## What Is the Scope API?
+
+The Scope API is Yantra's programmatic interface. Anything you can do in the Yantra app, you can automate through the API:
+
+- **Search** — Execute searches programmatically and retrieve structured results.
+- **Conversations** — Create and continue multi-turn AI conversations.
+- **Content management** — Upload, index, and manage documents.
+- **Export** — Retrieve search history, conversations, and analytics data.
+- **Webhooks** — Receive real-time notifications for events.
+
+---
+
+## Quick Start
+
+### 1. Get your API key
+
+1. Go to **Settings > API > API Keys**.
+2. Click **Generate New Key**.
+3. Copy the key — it's only shown once.
+
+### 2. Make your first request
+
+```bash
+curl -X POST https://api.yantra.ai/v1/search \
+  -H "Authorization: Bearer YOUR_API_KEY" \
+  -H "Content-Type: application/json" \
+  -d '{"query": "best practices for microservices architecture", "mode": "pro"}'
+```
+
+### 3. Parse the response
+
+```json
+{
+    "data": {
+        "answer": "Microservices architecture best practices include...",
+        "sources": [{ "title": "...", "url": "...", "snippet": "..." }]
+    },
+    "meta": {
+        "request_id": "req_abc123",
+        "model": "gpt-4o",
+        "latency_ms": 847
+    }
+}
+```
+
+---
+
+## Core Endpoints
+
+| Endpoint                         | Method | Description                               |
+| -------------------------------- | ------ | ----------------------------------------- |
+| `/v1/search`                     | POST   | Execute a search query                    |
+| `/v1/conversations`              | POST   | Start a new conversation                  |
+| `/v1/conversations/:id/messages` | POST   | Add a message to an existing conversation |
+| `/v1/content`                    | POST   | Upload and index a document               |
+| `/v1/content`                    | GET    | List indexed content                      |
+| `/v1/export`                     | GET    | Export user data                          |
+| `/v1/webhooks`                   | POST   | Register a webhook endpoint               |
+
+---
+
+## Authentication
+
+| Method    | Use case         | Header                            |
+| --------- | ---------------- | --------------------------------- |
+| API Key   | Server-to-server | `Authorization: Bearer <key>`     |
+| OAuth 2.0 | User-facing apps | Standard auth code flow with PKCE |
+
+---
+
+## Rate Limits
+
+| Plan       | Requests/minute | Concurrent requests |
+| ---------- | --------------- | ------------------- |
+| Free       | 20              | 2                   |
+| Pro        | 100             | 10                  |
+| Team       | 500             | 25                  |
+| Enterprise | Custom          | Custom              |
+
+When rate-limited, the API returns `429 Too Many Requests` with a `Retry-After` header.
+
+---
+
+## SDKs
+
+| Language   | Package       | Install                                  |
+| ---------- | ------------- | ---------------------------------------- |
+| Python     | `yantra-sdk`  | `pip install yantra-sdk`                 |
+| JavaScript | `@yantra/sdk` | `npm install @yantra/sdk`                |
+| Go         | `go-yantra`   | `go get github.com/yantra-app/go-yantra` |
+
+All SDKs include TypeScript types, automatic retry logic, and rate limit handling.
+
+---
+
+## Full Documentation
+
+For detailed endpoint documentation, request/response schemas, error codes, and examples, visit the [API Reference](https://docs.yantra.ai/api).

package/lib/templates/content/search-modes/deep-research.md

@@ -0,0 +1,109 @@
+# Deep Research
+
+Deep Research is Yantra's most comprehensive search mode — designed for complex topics that require thorough analysis across dozens of sources.
+
+---
+
+## What Deep Research Does
+
+When you activate Deep Research, Yantra:
+
+1. **Plans the research** — Breaks your question into sub-questions and identifies the key angles to explore.
+2. **Gathers sources** — Fetches and analyzes 20–50+ sources, including web pages, academic papers, and documents.
+3. **Synthesizes findings** — Cross-references information across all sources to build a comprehensive picture.
+4. **Generates a report** — Delivers a structured, citation-rich response organized by key findings.
+
+> **Time:** Deep Research takes 30 seconds to 2 minutes, depending on topic complexity. A progress indicator shows you what stage the research is in.
+
+---
+
+## When to Use Deep Research
+
+| Use case            | Example query                                                                            |
+| ------------------- | ---------------------------------------------------------------------------------------- |
+| Literature review   | "What does current research say about the effectiveness of remote work on productivity?" |
+| Market analysis     | "Compare the top 5 project management tools for engineering teams"                       |
+| Strategic planning  | "What are the risks and opportunities of expanding into the EU market in 2025?"          |
+| Technical deep dive | "Explain the trade-offs between microservices and monolithic architecture"               |
+| Competitive intel   | "How does Notion's AI strategy compare to Confluence's?"                                 |
+
+### When NOT to use it
+
+- Simple factual lookups ("What year was Python created?") → Use Regular search
+- Quick comparisons ("React vs Vue popularity") → Use Pro Search
+- Time-sensitive queries ("stock price right now") → Use Regular search with real-time mode
+
+---
+
+## Research Report Structure
+
+A typical Deep Research response includes:
+
+| Section               | Description                                                |
+| --------------------- | ---------------------------------------------------------- |
+| **Executive summary** | 2–3 sentence overview of key findings                      |
+| **Key findings**      | Numbered list of the most important insights               |
+| **Detailed analysis** | Organized prose with headers for each subtopic             |
+| **Comparison table**  | When applicable, a table comparing options or perspectives |
+| **Source list**       | 20–50+ cited sources at the bottom                         |
+
+---
+
+## Research Quality Features
+
+### Source diversity
+
+Deep Research deliberately gathers sources from multiple categories:
+
+- Academic papers and preprints
+- Industry reports and whitepapers
+- News articles and journalism
+- Official documentation and guides
+- Community discussions and expert opinions
+
+### Conflict resolution
+
+When sources disagree, Yantra:
+
+- Explicitly notes the disagreement
+- Presents both sides with their supporting evidence
+- Indicates which position has stronger or more recent support
+
+### Confidence indicators
+
+Each finding is tagged with a confidence level based on:
+
+- Number of corroborating sources
+- Source authority and credibility
+- Recency of the data
+
+---
+
+## Customizing Deep Research
+
+### Focus areas
+
+After clicking Deep Research, you can optionally narrow the scope:
+
+- **Academic only** — Limit to peer-reviewed and scholarly sources
+- **Recent only** — Focus on content published within a specified timeframe
+- **Specific domains** — Restrict to particular websites or source types
+
+### Export
+
+Deep Research reports are designed for export:
+
+- **PDF** — Formatted report with all citations
+- **Markdown** — For pasting into documentation or Notion
+- **Google Docs** — Direct export for collaborative editing
+
+---
+
+## Availability
+
+| Plan       | Deep Research access               |
+| ---------- | ---------------------------------- |
+| Free       | 3 per day                          |
+| Pro        | Unlimited                          |
+| Team       | Unlimited for all members          |
+| Enterprise | Unlimited with priority processing |