feat(obsidian): add gmail-digest + make paths configurable (v1.4.0)

plugins/obsidian/.claude-plugin/plugin.json

@@ -1,5 +1,5 @@
 {
   "name": "arevlo:obsidian",
-  "description": "Capture screenshots and context to Obsidian Zettelkasten vault as fragment notes. Save external links to note tables.",
-  "version": "1.2.1"
+  "description": "Capture screenshots and context to Obsidian Zettelkasten vault as fragment notes. Save external links to note tables. Generate Gmail digest summaries.",
+  "version": "1.4.0"
 }

plugins/obsidian/README.md

@@ -1,6 +1,6 @@
 # Obsidian Capture Plugin
 
-Capture screenshots and context directly to your Obsidian Zettelkasten vault as fragment notes.
+Capture screenshots and context directly to your Obsidian Zettelkasten vault as fragment notes. Save external links to note tables. Generate Gmail digest summaries.
 
 ## Commands
 
@@ -92,14 +92,54 @@ The command creates/updates a table in your note with this structure:
 | Constitutional AI Research Paper | #research #alignment #source/anthropic | [Constitutional AI Research Paper](https://anthropic.com/research/constitutional-ai) | Foundational paper on RLHF with AI feedback |
 ```
 
-## Configuration
+### `/arevlo:obsidian:gmail-digest`
+
+Open Gmail in Chrome, read all unread emails, categorize them, and save a bucketed digest to your Obsidian vault.
 
-The plugin expects your Obsidian vault at:
+**Usage:**
 ```
-/Users/arevlo/Library/Mobile Documents/com~apple~CloudDocs/zk
+/arevlo:obsidian:gmail-digest
 ```
 
-To change this, edit the `VAULT_PATH` in `commands/capture.md`.
+**What it does:**
+1. Opens Gmail in Chrome filtered to unread emails
+2. Reads each unread email (opening them marks as read)
+3. Categorizes emails into buckets:
+   - **Action Required** — needs a response or task
+   - **Calendar & Scheduling** — meeting invites, schedule changes
+   - **Conversations** — ongoing threads/replies
+   - **FYI / Informational** — newsletters, notifications, updates
+   - **Automated / System** — receipts, alerts, service notifications
+4. Generates a structured markdown digest
+5. Saves to `{digest_output_path}/gmail-digest-YYYY-MM-DD.md`
+6. Shows summary of processed emails and categories
+
+**Example output file:** `gmail-digest-2026-02-07.md` (saved to your configured `digest_output_path`)
+
+**Notes:**
+- Running multiple times per day appends a counter (e.g., `gmail-digest-2026-02-07-2.md`)
+- Requires Chrome with the Claude in Chrome extension
+- Emails are marked as read by being opened during processing
+
+### `/arevlo:obsidian:configure`
+
+Set up or update the Obsidian plugin configuration (vault path, digest output path).
+
+**Usage:**
+```
+/arevlo:obsidian:configure
+```
+
+## Configuration
+
+On first use, any command will prompt you to configure your paths. Configuration is stored in `~/.claude/obsidian-plugin.json`:
+
+| Field | Used By | Description |
+|-------|---------|-------------|
+| `vault_path` | `/capture`, `/save-link` | Absolute path to your Obsidian vault |
+| `digest_output_path` | `/gmail-digest` | Absolute path for Gmail digest files |
+
+To reconfigure, run `/arevlo:obsidian:configure` or edit `~/.claude/obsidian-plugin.json` directly.
 
 ## Installation
 
@@ -147,5 +187,7 @@ Use `mcp__obsidian-zettelkasten__process_fragment` to convert fragments to primi
 
 ## Version History
 
+- **1.4.0** - Made paths configurable via `~/.claude/obsidian-plugin.json`; added `/arevlo:obsidian:configure` command
+- **1.3.0** - Added `/arevlo:obsidian:gmail-digest` command for Gmail digest summaries
 - **1.2.0** - Added `/arevlo:obsidian:save-link` command for saving external links to notes
 - **1.0.0** - Initial release with screenshot capture to fragment notes

plugins/obsidian/commands/capture.md

@@ -21,10 +21,31 @@ When you share a screenshot (from Slack, Figma, browser, etc.) along with option
 
 ## Configuration
 
-**Obsidian vault path:** `/Users/arevlo/Library/Mobile Documents/com~apple~CloudDocs/zk`
+Configuration is stored in `~/.claude/obsidian-plugin.json`. The `vault_path` field is used for fragment captures.
 
 ## Steps
 
+### 0. Load Configuration
+
+Before doing anything else, read the configuration file:
+
+1. Use the `Read` tool to read `~/.claude/obsidian-plugin.json`
+2. **If the file does not exist:**
+   - Inform the user: "No Obsidian plugin configuration found. Let's set it up."
+   - Use `AskUserQuestion` to ask for their Obsidian vault path (absolute path, no `~`)
+   - Use `AskUserQuestion` to ask for their digest output directory (absolute path, no `~`)
+   - Use `Bash` to create the config file:
+     ```bash
+     mkdir -p ~/.claude && cat > ~/.claude/obsidian-plugin.json << 'ENDCONFIG'
+     {
+       "vault_path": "{user's vault path}",
+       "digest_output_path": "{user's digest path}"
+     }
+     ENDCONFIG
+     ```
+   - Continue with the values provided
+3. **If the file exists:** Parse the JSON and extract `vault_path` for use in subsequent steps
+
 ### 1. Gather Information First
 
 **Before doing any file operations**, collect all the information needed:
@@ -106,8 +127,8 @@ echo "━━━━━━━━━━━━━━━━━━━━━━━━
 Then continue with the copy operation:
 
 ```bash
-# Set vault path and fragment folder
-VAULT_PATH="/Users/arevlo/Library/Mobile Documents/com~apple~CloudDocs/zk"
+# Set vault path from config (loaded in Step 0) and fragment folder
+VAULT_PATH="{vault_path from config}"
 FRAGMENT_FOLDER="{category}/fragments"
 
 # Create _attachments folder if needed

plugins/obsidian/commands/configure.md

@@ -0,0 +1,57 @@
+---
+description: Configure Obsidian plugin vault paths
+allowed-tools: Read,Write,AskUserQuestion
+---
+
+Configure or update the Obsidian plugin paths used by `/capture`, `/save-link`, and `/gmail-digest`.
+
+## What this does
+
+Sets up `~/.claude/obsidian-plugin.json` with your personal vault and output paths so all Obsidian plugin commands work correctly.
+
+## Steps
+
+### 1. Check for Existing Configuration
+
+Use the `Read` tool to read `~/.claude/obsidian-plugin.json`.
+
+- **If the file exists:** Show the current configuration values to the user.
+- **If the file does not exist:** Inform the user that no configuration exists yet.
+
+### 2. Ask for Vault Path
+
+Use `AskUserQuestion` to ask:
+- Question: "What is the absolute path to your Obsidian vault?"
+- If updating, show the current value as context in the question
+- Remind the user to use an absolute path (no `~`)
+
+### 3. Ask for Digest Output Path
+
+Use `AskUserQuestion` to ask:
+- Question: "What is the absolute path for Gmail digest output files?"
+- If updating, show the current value as context in the question
+- Remind the user to use an absolute path (no `~`)
+
+### 4. Save Configuration
+
+Use the `Write` tool to save the config to `~/.claude/obsidian-plugin.json`:
+
+```json
+{
+  "vault_path": "{user's vault path}",
+  "digest_output_path": "{user's digest path}"
+}
+```
+
+### 5. Confirm
+
+Show the saved configuration:
+
+```
+Obsidian plugin configured!
+
+  vault_path:         {vault_path}
+  digest_output_path: {digest_output_path}
+
+Config saved to: ~/.claude/obsidian-plugin.json
+```

plugins/obsidian/commands/gmail-digest.md

@@ -0,0 +1,207 @@
+---
+description: Analyze unread Gmail emails and create a bucketed digest in Obsidian vault
+allowed-tools: mcp__claude-in-chrome__tabs_context_mcp,mcp__claude-in-chrome__tabs_create_mcp,mcp__claude-in-chrome__navigate,mcp__claude-in-chrome__read_page,mcp__claude-in-chrome__javascript_tool,mcp__claude-in-chrome__computer,Write,Bash,AskUserQuestion,Read
+---
+
+Open Gmail in Chrome, read all unread emails, categorize them into bucketed summaries, save as a dated markdown file in the Obsidian vault, and mark emails as read.
+
+## Configuration
+
+Configuration is stored in `~/.claude/obsidian-plugin.json`. The `digest_output_path` field is used for the output directory.
+
+**Output filename:** `gmail-digest-YYYY-MM-DD.md` (appends counter if file exists: `gmail-digest-2026-02-07-2.md`)
+
+## Steps
+
+### 0. Load Configuration
+
+Before doing anything else, read the configuration file:
+
+1. Use the `Read` tool to read `~/.claude/obsidian-plugin.json`
+2. **If the file does not exist:**
+   - Inform the user: "No Obsidian plugin configuration found. Let's set it up."
+   - Use `AskUserQuestion` to ask for their Obsidian vault path (absolute path, no `~`)
+   - Use `AskUserQuestion` to ask for their digest output directory (absolute path, no `~`)
+   - Use `Bash` to create the config file:
+     ```bash
+     mkdir -p ~/.claude && cat > ~/.claude/obsidian-plugin.json << 'ENDCONFIG'
+     {
+       "vault_path": "{user's vault path}",
+       "digest_output_path": "{user's digest path}"
+     }
+     ENDCONFIG
+     ```
+   - Continue with the values provided
+3. **If the file exists:** Parse the JSON and extract `digest_output_path` for use in subsequent steps
+
+### 1. Open Gmail Unread View
+
+Use `mcp__claude-in-chrome__tabs_create_mcp` to create a new tab, then `mcp__claude-in-chrome__navigate` to:
+
+```
+https://mail.google.com/mail/u/0/#search/is%3Aunread
+```
+
+Wait for the page to load fully.
+
+### 2. Read the Unread Email List
+
+Use `mcp__claude-in-chrome__read_page` to extract the list of unread emails visible on the page. Capture:
+- Sender name
+- Subject line
+- Snippet/preview text
+- Approximate time received
+
+If there are **no unread emails**, skip to Step 7 and inform the user:
+```
+No unread emails found. Nothing to digest!
+```
+
+### 3. Deep Read Each Email
+
+For each unread email in the list:
+
+1. **Click into the email** using `mcp__claude-in-chrome__computer` (click on the email row). Opening an email in Gmail **automatically marks it as read**.
+2. **Read the full content** using `mcp__claude-in-chrome__read_page` to get the complete email body, sender details, and any attachments info.
+3. **Navigate back** to the unread list. Use `mcp__claude-in-chrome__navigate` to go back to `https://mail.google.com/mail/u/0/#search/is%3Aunread` (the list will now show remaining unread emails).
+4. **Repeat** until all unread emails have been read.
+
+**Pagination:** If after reading all visible emails the unread list still shows more emails (pagination), continue reading those as well.
+
+**Important:** Keep a running list of all email data collected. For each email, store:
+- Sender (name and email)
+- Subject
+- Date/time
+- Full body content (or meaningful summary for very long emails)
+- Whether it's part of a thread
+- Any action items or requests mentioned
+
+### 4. Analyze & Categorize
+
+Bucket every email into exactly one of these categories:
+
+- **Action Required** — Emails needing a response, decision, or task from the user
+- **Calendar & Scheduling** — Meeting invites, schedule changes, RSVPs
+- **Conversations** — Ongoing threads, replies, discussions
+- **FYI / Informational** — Newsletters, announcements, product updates, notifications
+- **Automated / System** — Receipts, shipping confirmations, alerts, service notifications, automated reports
+
+For each email, determine:
+- **Category** (one of the above)
+- **Key takeaway** (1-2 sentence summary of what matters)
+- **Urgency** (if applicable, especially for Action Required items)
+
+### 5. Generate Markdown
+
+Build the digest markdown file with this structure:
+
+```markdown
+---
+date: YYYY-MM-DD
+type: gmail-digest
+emails_processed: {count}
+---
+
+# Gmail Digest — {Mon DD, YYYY}
+
+> {count} unread emails processed and marked as read.
+
+## Action Required ({count})
+
+### {Subject} — {Sender Name}
+{Key takeaway / summary. Include urgency or deadlines if mentioned.}
+
+...
+
+## Calendar & Scheduling ({count})
+
+### {Subject} — {Sender Name}
+{Key takeaway / summary.}
+
+...
+
+## Conversations ({count})
+
+### {Subject} — {Sender Name}
+{Key takeaway / summary. Note thread context if relevant.}
+
+...
+
+## FYI / Informational ({count})
+
+### {Subject} — {Sender Name}
+{Key takeaway / summary.}
+
+...
+
+## Automated / System ({count})
+
+### {Subject} — {Sender Name}
+{Key takeaway / summary.}
+
+...
+```
+
+**Rules:**
+- Only include category sections that have emails. Skip empty categories entirely.
+- Order categories by importance: Action Required first, Automated/System last.
+- Each email entry should be concise — 1-3 sentences max for the summary.
+- Use the sender's display name, not their email address.
+- Format dates in the header as "Feb 7, 2026" style.
+
+### 6. Save the File
+
+Determine the filename:
+
+```bash
+DATE=$(date +%Y-%m-%d)
+OUTPUT_DIR="{digest_output_path from config}"
+BASE="${OUTPUT_DIR}/gmail-digest-${DATE}"
+FILE="${BASE}.md"
+
+# If file already exists, append a counter
+if [ -f "$FILE" ]; then
+  COUNTER=2
+  while [ -f "${BASE}-${COUNTER}.md" ]; do
+    COUNTER=$((COUNTER + 1))
+  done
+  FILE="${BASE}-${COUNTER}.md"
+fi
+
+echo "$FILE"
+```
+
+Use the `Write` tool to save the markdown content to the resolved file path.
+
+### 7. Confirm to User
+
+**If no unread emails were found** (skipped from Step 2), just inform the user:
+
+```
+No unread emails found. Nothing to digest!
+```
+
+**If emails were processed**, show a summary:
+
+```
+Gmail Digest Complete
+
+  Emails processed: {count}
+  Categories:
+    - Action Required: {n}
+    - Calendar & Scheduling: {n}
+    - Conversations: {n}
+    - FYI / Informational: {n}
+    - Automated / System: {n}
+
+  Saved to: {digest_output_path}/gmail-digest-YYYY-MM-DD.md
+  All emails marked as read.
+```
+
+## Notes
+
+- **Reading = marking as read:** Opening each email in Gmail's web UI automatically marks it as read. No additional action is needed to mark emails as read.
+- **Long emails:** For very long emails (marketing newsletters, legal notices), summarize the key point rather than capturing everything.
+- **Attachments:** Note if an email has attachments but don't attempt to download them.
+- **Errors:** If Gmail fails to load or Chrome is unresponsive, inform the user and suggest they check their browser/internet connection.
+- **Multiple runs per day:** The filename counter logic handles this gracefully.