Merge pull request #40 from eddiesanjuan/feature/v2.5.0-announcement

feat: v2.5.0 announcement page and social content

Author: eddiesanjuan

launch-content/ONE-CLICK-KIT.md

@@ -1,7 +1,7 @@
 ---
 title: I Built an MCP Server That Gives AI Agents Eyes and Ears
 published: false
-description: markupr records your screen and voice, then produces structured Markdown with screenshots placed at the exact moments that matter. Now your AI coding agent can see what you see.
+description: markupr records your screen and voice, structures feedback with screenshots, and pushes it directly to GitHub Issues or Linear. Your AI coding agent sees what you see -- and files the ticket.
 tags: ai, opensource, webdev, productivity
 cover_image:
 ---
@@ -123,9 +123,9 @@ npx markupr analyze ./recording.mov --output ./reports --verbose
 
 No Electron, no desktop app. Works in CI/CD pipelines. An AI agent can run this command to process a recording programmatically.
 
-### 3. MCP Server (New in v2.4.0)
+### 3. MCP Server
 
-This is the one I'm most excited about. MCP (Model Context Protocol) lets AI coding agents call tools. The markupr MCP server gives your agent 6 tools for screen capture and recording.
+MCP (Model Context Protocol) lets AI coding agents call tools. The markupr MCP server gives your agent 6 tools for screen capture and recording.
 
 #### Setup
 
@@ -176,6 +176,56 @@ No screenshotting. No uploading. No describing pixel positions. The agent looks
 
 For longer reviews, the agent can call `capture_with_voice({ duration: 60 })` while you narrate issues for a minute. It gets back a full structured report with every issue you described, each with the corresponding screenshot.
 
+## New in v2.5.0: Feedback Delivery Pipeline
+
+v2.5.0 closes the loop. markupr doesn't just capture and structure feedback -- it delivers it to your issue tracker.
+
+### Push to GitHub Issues
+
+```bash
+markupr push github --repo owner/repo
+```
+
+Creates a GitHub issue directly from your session. Screenshots are uploaded and embedded inline. The issue body is structured Markdown that humans and AI agents can both parse.
+
+### Push to Linear
+
+```bash
+markupr push linear --team KEY
+```
+
+Creates a Linear issue with full context -- transcript, screenshots, severity labels. Record the bug, push to the backlog, done.
+
+### Output Templates
+
+```bash
+# Format output for your issue tracker
+npx markupr analyze ./recording.mov --template github-issue
+npx markupr analyze ./recording.mov --template linear
+npx markupr analyze ./recording.mov --template jira
+npx markupr analyze ./recording.mov --template json
+```
+
+Five template formats: standard Markdown (default), GitHub Issue, Linear, Jira, and JSON. Pick the format your tools consume.
+
+### Watch Mode
+
+```bash
+markupr watch ./recordings
+```
+
+Monitors a directory and auto-processes any new recording that appears. Drop a `.mov` in the folder, get a structured report back. Pairs well with automated screen recording tools.
+
+### GitHub Action
+
+```yaml
+- uses: eddiesanjuan/markupr-action@v1
+```
+
+Runs markupr in CI/CD. Push a commit, the action analyzes visual changes and posts structured feedback as a PR comment with screenshots. Teams get automated visual QA without manual recording.
+
+The pipeline went from "record and structure" to "record, structure, and deliver." The feedback loop is closed.
+
 ## Sample Output
 
 Here's what a processed session looks like:
@@ -245,16 +295,18 @@ You control when and whether data leaves your machine.
 - Linux support for the desktop app
 - Browser extension for capturing web app sessions
 - Collaborative sessions (multiple narrators, one recording)
-- Plugin system for custom output formats
+- Jira Cloud push integration (currently template-only)
+- Slack/Discord notification hooks
 
 ## Links
 
 - **GitHub**: [github.com/eddiesanjuan/markupr](https://github.com/eddiesanjuan/markupr)
 - **Site**: [markupr.com](https://markupr.com)
 - **MCP server**: `npx markupr-mcp` (zero install)
 - **CLI**: `npx markupr analyze ./video.mov`
+- **GitHub Action**: `eddiesanjuan/markupr-action@v1`
 - **npm**: [npmjs.com/package/markupr](https://www.npmjs.com/package/markupr)
 
-Open source, MIT licensed. 644 tests across 39 files. Contributions welcome.
+Open source, MIT licensed. 860 tests across 44 files. Contributions welcome.
 
 If markupr saves you time, consider [supporting development on Ko-fi](https://ko-fi.com/eddiesanjuan).

launch-content/devto-article.md

@@ -27,7 +27,7 @@ There are three ways to use it:
 
 **CLI** -- `npx markupr analyze ./recording.mov` processes any screen recording into a Markdown report with extracted frames. No Electron, no desktop app needed.
 
-**MCP server** -- This is the v2.4.0 addition. Add 3 lines of JSON to your Claude Code or Cursor config and your agent gets 6 tools: screenshot capture, screen+voice recording, video analysis, and interactive recording sessions. The agent can see your screen mid-conversation.
+**MCP server** -- Add 3 lines of JSON to your Claude Code or Cursor config and your agent gets 6 tools: screenshot capture, screen+voice recording, video analysis, and interactive recording sessions. The agent can see your screen mid-conversation.
 
 ```json
 {
@@ -39,10 +39,21 @@ There are three ways to use it:
 
 After setup, you can say "the sidebar is broken on mobile, can you see it?" and the agent captures a screenshot, sees the issue, and fixes it. No copy-pasting.
 
+**New in v2.5.0: markupr now delivers feedback to your issue tracker.**
+
+- **Output Templates** -- `--template github-issue` or `--template linear` formats your feedback as structured issues, not just markdown. Also supports JSON and Jira formats.
+- **Push to GitHub Issues** -- `markupr push github --repo owner/repo` creates a GitHub issue directly from a session. Screenshots are uploaded and embedded.
+- **Push to Linear** -- `markupr push linear --team KEY` creates a Linear issue with full context. Record the bug, push to the backlog, done.
+- **Watch Mode** -- `markupr watch ./dir` monitors a directory and auto-processes any new recording that appears. Drop a screen recording in a folder, get a structured report back.
+- **GitHub Action** -- `eddiesanjuan/markupr-action@v1` runs markupr in CI. Push a commit, the action analyzes visual changes, and posts structured feedback as a PR comment. Teams get automated visual QA without manual recording.
+
+The pipeline went from "record and structure" to "record, structure, and deliver." You record the bug, markupr writes it up, and pushes it to your issue tracker. The feedback loop is closed.
+
 Everything runs locally by default. Whisper transcription happens on your machine. No telemetry, no tracking, no accounts. The only external calls happen if you explicitly configure an OpenAI key for cloud transcription or an Anthropic key for AI-enhanced analysis.
 
-Open source, MIT licensed. 644 tests.
+Open source, MIT licensed. 860 tests.
 
 Repo: https://github.com/eddiesanjuan/markupr
 Site: https://markupr.com
 npm: `npx markupr-mcp` (MCP server) / `npx markupr analyze` (CLI)
+GitHub Action: `eddiesanjuan/markupr-action@v1`

launch-content/hackernews.md

@@ -2,7 +2,7 @@
 
 > **Package:** `markupr` on npm (includes `markupr-mcp` binary)
 > **Repo:** https://github.com/eddiesanjuan/markupr
-> **Version:** 2.4.0
+> **Version:** 2.5.0
 > **Transport:** stdio
 > **Tools:** capture_screenshot, capture_with_voice, analyze_video, analyze_screenshot, start_recording, stop_recording
 
@@ -44,17 +44,17 @@ curl -L "https://github.com/modelcontextprotocol/registry/releases/latest/downlo
 {
   "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
   "name": "io.github.eddiesanjuan/markupr",
-  "description": "Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice narration, analyze video, and generate structured feedback documents.",
+  "description": "Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice narration, analyze video, generate structured feedback, and push to GitHub Issues or Linear.",
   "repository": {
     "url": "https://github.com/eddiesanjuan/markupr",
     "source": "github"
   },
-  "version": "2.4.0",
+  "version": "2.5.0",
   "packages": [
     {
       "registryType": "npm",
       "identifier": "markupr",
-      "version": "2.4.0",
+      "version": "2.5.0",
       "transport": {
         "type": "stdio"
       },
@@ -88,7 +88,7 @@ curl "https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.edd
 Add under the **Developer Tools** section, maintaining alphabetical order:
 
 ```markdown
-- [markupr](https://github.com/eddiesanjuan/markupr) 📇 🏠 🍎 🪟 - Screen capture and voice recording MCP server for AI coding agents. Capture screenshots with voice narration, analyze video, and generate structured Markdown feedback documents.
+- [markupr](https://github.com/eddiesanjuan/markupr) 📇 🏠 🍎 🪟 - Screen capture and voice recording MCP server for AI coding agents. Capture screenshots with voice narration, analyze video, generate structured Markdown feedback, and push to GitHub Issues or Linear. Includes output templates, watch mode, and GitHub Action.
 ```
 
 ### PR Instructions
@@ -106,7 +106,7 @@ Add under the **Developer Tools** section, maintaining alphabetical order:
 - **Language:** TypeScript
 - **Scope:** Local (runs on device)
 - **OS:** macOS, Windows
-- **Description:** Screen capture and voice recording MCP server for AI coding agents. 6 tools: capture_screenshot, capture_with_voice, analyze_video, analyze_screenshot, start_recording, stop_recording.
+- **Description:** Screen capture and voice recording MCP server for AI coding agents. 6 tools: capture_screenshot, capture_with_voice, analyze_video, analyze_screenshot, start_recording, stop_recording. v2.5.0 adds push to GitHub Issues/Linear, output templates, watch mode, and a GitHub Action.
 - **Category:** Developer Tools
 ```
 
@@ -122,7 +122,7 @@ Add under the **Developer Tools** section, maintaining alphabetical order:
 Add under the **Development Tools** section:
 
 ```markdown
-- **[markupr](https://github.com/eddiesanjuan/markupr)** - Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice, analyze video, and generate structured feedback.
+- **[markupr](https://github.com/eddiesanjuan/markupr)** - Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice, analyze video, generate structured feedback, and push to GitHub Issues or Linear.
 ```
 
 ---
@@ -174,7 +174,7 @@ If not indexed within a week, join their Discord to request manual addition.
 
 ### Form Fields
 - **Server Name:** markupr
-- **Short Description:** Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice narration, analyze video, and generate structured Markdown feedback documents.
+- **Short Description:** Screen capture and voice recording MCP server for AI coding agents. Capture screenshots, record voice narration, analyze video, generate structured Markdown feedback, and push to GitHub Issues or Linear.
 - **Link:** https://github.com/eddiesanjuan/markupr
 - **Category:** development
 - **Contact Email:** eddie@markupr.com
@@ -206,7 +206,7 @@ If not indexed within a week, join their Discord to request manual addition.
 **Server Name:** markupr
 **GitHub:** https://github.com/eddiesanjuan/markupr
 **npm:** markupr (run via `npx markupr-mcp`)
-**Description:** Screen capture and voice recording MCP server for AI coding agents. Capture screenshots with voice narration, analyze video, and generate structured Markdown feedback documents for code review.
+**Description:** Screen capture and voice recording MCP server for AI coding agents. Capture screenshots with voice narration, analyze video, generate structured Markdown feedback, and push to GitHub Issues or Linear. v2.5.0 includes output templates, watch mode, and GitHub Action.
 **Tools:** capture_screenshot, capture_with_voice, analyze_video, analyze_screenshot, start_recording, stop_recording
 **Category:** Developer Tools
 ```

launch-content/mcp-submissions.md

@@ -22,9 +22,9 @@ Everything needed to launch markupr on Product Hunt. Copy-paste ready.
 
 **Description** (260 chars max):
 
-> markupr records your screen while you talk through bugs. An intelligent pipeline transcribes your voice, extracts frames at the exact moments you described, and produces structured Markdown your AI coding agent can act on. Open source. Free. Works with Claude Code, Cursor, and any MCP client.
+> markupr records your screen, structures feedback with screenshots at the moments that matter, and pushes it to GitHub Issues or Linear. CLI, MCP server, GitHub Action. Open source. Free. Works with Claude Code, Cursor, and any MCP client.
 
-(258 chars)
+(240 chars)
 
 **Topics:**
 - Developer Tools
@@ -51,14 +51,25 @@ Everything needed to launch markupr on Product Hunt. Copy-paste ready.
 >
 > The output is purpose-built for AI coding agents. Paste the file path into Claude Code, Cursor, or Windsurf and the agent has everything it needs to start fixing.
 >
-> What I'm most excited about in v2.4.0 is the **MCP server**. Your AI agent can now see your screen and hear your voice mid-conversation. Three lines of JSON config, zero install (`npx markupr-mcp`), and suddenly your agent has `capture_screenshot`, `capture_with_voice`, `start_recording`, and `stop_recording` tools. The agent looks at your screen and acts. No copy-pasting screenshots. No describing layouts in text.
+> The **MCP server** gives your AI agent eyes and ears. Three lines of JSON config, zero install (`npx markupr-mcp`), and your agent has `capture_screenshot`, `capture_with_voice`, `start_recording`, and `stop_recording` tools. The agent looks at your screen and acts.
+>
+> What I'm most excited about in **v2.5.0** is the delivery pipeline. markupr now pushes feedback directly to your issue tracker:
+> - `markupr push github --repo owner/repo` -- creates a GitHub issue with embedded screenshots
+> - `markupr push linear --team KEY` -- creates a Linear issue with full context
+> - `markupr watch ./dir` -- auto-process recordings from a directory
+> - `--template github-issue/linear/jira/json` -- output in the format your tools consume
+> - **GitHub Action** (`eddiesanjuan/markupr-action@v1`) -- automated visual QA in CI/CD
+>
+> The pipeline went from "record and structure" to "record, structure, and deliver." You see a bug, record it, and it lands in your backlog. The feedback loop is closed.
 >
 > markupr is fully open source and MIT licensed. No telemetry, no tracking, no analytics. Local Whisper transcription runs entirely on your machine -- no API key needed to start. If you want cloud transcription or AI-enhanced analysis, bring your own OpenAI or Anthropic keys.
 >
-> Three ways to use it:
+> Five ways to use it:
 > 1. **Desktop app** -- menu bar on macOS, system tray on Windows
 > 2. **CLI** -- `npx markupr analyze ./recording.mov`
 > 3. **MCP server** -- `npx markupr-mcp` in your IDE config
+> 4. **Watch Mode** -- `markupr watch ./dir` auto-processes recordings
+> 5. **GitHub Action** -- `eddiesanjuan/markupr-action@v1` in CI/CD
 >
 > I'd love your honest feedback. What would make this more useful in your workflow? What's missing? I'm shipping fast and building in public.
 >
@@ -82,7 +93,9 @@ Everything needed to launch markupr on Product Hunt. Copy-paste ready.
 >
 > The MCP server is the one I think will surprise people. Zero install -- npx handles everything. Add 3 lines to your Claude Code or Cursor config and your agent can look at your screen mid-conversation. It's the bridge between "I can see the bug" and "my agent can fix it."
 >
-> **644 tests** across 39 files. MIT licensed. No vendor lock-in. The whole codebase is on GitHub.
+> v2.5.0 adds a delivery layer: push structured feedback directly to GitHub Issues (`markupr push github --repo owner/repo`) or Linear (`markupr push linear --team KEY`). There's a template system (`--template github-issue`, `--template linear`, `--template jira`, `--template json`), a watch mode for auto-processing recordings, and a GitHub Action (`eddiesanjuan/markupr-action@v1`) for visual QA in CI/CD.
+>
+> **860 tests** across 44 files. MIT licensed. No vendor lock-in. The whole codebase is on GitHub.
 
 ---
 
@@ -175,9 +188,9 @@ Send this via DM, email, or group chat. Keep it personal -- never mass-blast.
 
 > I just launched markupr on Product Hunt. It's an open-source dev tool I've been building to solve a specific problem: describing visual bugs to AI coding agents.
 >
-> The workflow: press a hotkey, narrate what you see on screen, press the hotkey again. markupr transcribes your voice, extracts video frames at the exact timestamps from your narration, and produces structured Markdown your AI agent can consume directly. New in v2.4.0: an MCP server that gives your agent capture_screenshot and capture_with_voice tools -- zero install, just `npx markupr-mcp`.
+> The workflow: press a hotkey, narrate what you see on screen, press the hotkey again. markupr transcribes your voice, extracts video frames at the exact timestamps from your narration, and produces structured Markdown your AI agent can consume directly. New in v2.5.0: push feedback straight to GitHub Issues or Linear, output templates for different trackers, watch mode for auto-processing, and a GitHub Action for CI/CD visual QA.
 >
-> Free, MIT licensed, no telemetry. Desktop app + CLI + MCP server.
+> Free, MIT licensed, no telemetry. Desktop app + CLI + MCP server + GitHub Action.
 >
 > Check it out: [PH link]
 > GitHub: https://github.com/eddiesanjuan/markupr