reeboot 1.0.0 → 1.3.0

package/skills/reeboot-tasks/SKILL.md

@@ -0,0 +1,98 @@
+---
+name: reeboot-tasks
+description: Manage the reeboot agent's own scheduled tasks — create, list, pause, resume, and cancel scheduled jobs. Use when setting up recurring tasks, one-time reminders, or managing the agent's task schedule.
+---
+
+# Reeboot Tasks
+
+Meta-skill for the agent to manage its own scheduler. Uses the built-in scheduler tools registered by the scheduler-tool extension.
+
+## Setup
+
+No external dependencies. The scheduler is built into reeboot.
+
+The following tools are available when the scheduler-tool extension is enabled:
+- `schedule_task` — create a new scheduled task
+- `list_tasks` — list all scheduled tasks
+- `cancel_task` — cancel a scheduled task by ID
+- `pause_task` — pause a scheduled task (stops it from running without deleting)
+- `resume_task` — resume a paused task
+
+Verify tools are available by asking: "What tools do you have access to?" — the scheduler tools should appear in the list.
+
+## Usage
+
+### Schedule formats
+
+**Cron expressions** (standard 5-field cron):
+```
+0 9 * * 1-5    # Every weekday at 9 AM
+0 */2 * * *    # Every 2 hours
+30 8 * * 1     # Every Monday at 8:30 AM
+0 0 1 * *      # First day of each month at midnight
+```
+
+**Interval expressions**:
+```
+every 30m      # Every 30 minutes
+every 2h       # Every 2 hours
+every 1d       # Every day
+every 15m      # Every 15 minutes
+```
+
+**One-time expressions**:
+```
+in 1h          # In 1 hour from now
+in 30m         # In 30 minutes
+in 2d          # In 2 days
+```
+
+### Tool usage
+
+```
+# Schedule a recurring task
+schedule_task({
+  name: "daily-standup-reminder",
+  schedule: "0 9 * * 1-5",
+  prompt: "Send a standup reminder to the team on WhatsApp"
+})
+
+# Schedule a one-time reminder
+schedule_task({
+  name: "meeting-reminder",
+  schedule: "in 45m",
+  prompt: "Remind me that the client call starts in 15 minutes"
+})
+
+# List all tasks
+list_tasks()
+
+# Pause a task
+pause_task({ id: "task-id-here" })
+
+# Resume a paused task
+resume_task({ id: "task-id-here" })
+
+# Cancel a task permanently
+cancel_task({ id: "task-id-here" })
+```
+
+### Example flows
+
+```
+User: "Remind me to check my email every weekday morning at 8 AM"
+→ schedule_task({
+    name: "morning-email-check",
+    schedule: "0 8 * * 1-5",
+    prompt: "Remind the user to check their email"
+  })
+→ "Done — I'll remind you every weekday at 8 AM."
+
+User: "What tasks do I have scheduled?"
+→ list_tasks()
+→ Present results in a readable format
+
+User: "Cancel the standup reminder"
+→ list_tasks() to find the ID
+→ cancel_task({ id: "<id>" })
+```

package/skills/send-message/SKILL.md

@@ -1,27 +1,65 @@
 ---
 name: send-message
-description: Send a message to a contact via WhatsApp or Signal. Use when the user asks you to send a message to someone via a messaging channel.
+description: Send a message back to the originating channel or to a specific contact via WhatsApp or Signal. Use when the user asks you to send, forward, or relay a message to someone via a messaging channel.
 ---
 
-# Send Message Skill
+# Send Message
 
-Use this skill when the user wants to send a message to a contact via a messaging channel (WhatsApp, Signal).
+Teaches the agent to send a message back to the originating channel or to a specific peer via reeboot's channel routing system.
 
-## Instructions
+## Setup
 
-1. Identify the target channel (WhatsApp or Signal) and recipient
-2. Confirm the message content with the user if not explicitly stated
-3. Use the appropriate channel tool to send the message
-4. Report success or failure back to the user
+No external dependencies. The send-message capability is built into reeboot's channel system.
 
-## When to Use
+Supported channels (must be configured in `~/.reeboot/config.json`):
+- **WhatsApp** — requires WhatsApp session to be active (`reeboot channels login whatsapp`)
+- **Signal** — requires Signal container running (`reeboot channels login signal`)
 
-- "Send a WhatsApp message to Alice saying..."
-- "Text Bob on Signal that..."
-- "Forward this to my team on WhatsApp"
+Verify channel status:
+```
+reeboot channels list
+```
+
+## Usage
+
+### Send a reply in the same turn
+
+Simply include your reply text as the response — reeboot routes it back to the originating channel automatically.
+
+### Send to a specific contact
+
+Use the `send_message` tool (registered by the channel routing system):
+
+```
+send_message({
+  channel: "whatsapp",       // or "signal"
+  peerId: "+1234567890",     // phone number or WhatsApp JID
+  content: "Hello from reeboot!"
+})
+```
+
+### When to use vs. just responding
+
+- **Just respond** — if you want to reply to the person who sent you the message (same turn, same channel). This is the default.
+- **Use send_message tool** — if you need to send to a *different* contact or channel, or send *proactively* (e.g., from a scheduled task).
+
+### Example flows
+
+```
+User (WhatsApp): "Text my wife that I'll be home late"
+→ Identify contact from user's address book or ask for phone number
+→ Confirm: "Send 'I'll be home late' to +1XXXXXXXXXX on WhatsApp?"
+→ send_message({ channel: "whatsapp", peerId: "+1XXXXXXXXXX", content: "I'll be home late" })
+→ Report: "Message sent."
+```
+
+```
+Scheduled task: daily standup reminder
+→ send_message({ channel: "whatsapp", peerId: "team@g.us", content: "Daily standup in 15 minutes!" })
+```
 
 ## Important
 
-- Always confirm the recipient and message content before sending
-- Do not send messages without explicit user instruction
+- Always confirm recipient and content before sending
+- Never send without explicit user instruction (or a pre-approved schedule)
 - Report delivery status or errors clearly

package/skills/slack/SKILL.md

@@ -0,0 +1,74 @@
+---
+name: slack
+description: Slack workspace operations via SLACK_BOT_TOKEN and curl — send messages, list channels, post to threads, upload files. Use when sending Slack messages, reading channels, or interacting with a Slack workspace.
+---
+
+# Slack
+
+Uses `SLACK_BOT_TOKEN` env var + curl against the Slack Web API to send messages, list channels, reply to threads, and upload files.
+
+## Setup
+
+1. Create a Slack App at https://api.slack.com/apps:
+   - Click **Create New App** → **From scratch**
+   - Give it a name and select your workspace
+
+2. Configure Bot Token Scopes under **OAuth & Permissions** → **Bot Token Scopes**:
+   - `channels:read` — list public channels
+   - `chat:write` — post messages
+   - `files:write` — upload files
+   - `groups:read` — list private channels (optional)
+   - `im:read` / `im:write` — direct messages (optional)
+
+3. Install to workspace:
+   - Click **Install to Workspace** → **Allow**
+   - Copy the **Bot User OAuth Token** (starts with `xoxb-`)
+
+4. Set the environment variable:
+   ```
+   export SLACK_BOT_TOKEN=xoxb-xxxxxxxxxxxx-xxxxxxxxxxxx-xxxxxxxxxxxxxxxxxxxxxxxx
+   ```
+   Add to your shell profile for persistence.
+
+5. Invite the bot to channels:
+   ```
+   /invite @your-bot-name
+   ```
+
+6. Verify:
+   ```
+   curl -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+        https://slack.com/api/auth.test
+   ```
+
+## Usage
+
+```bash
+# List channels
+curl -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+     "https://slack.com/api/conversations.list?limit=50"
+
+# Post a message
+curl -X POST \
+  -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+  -H "Content-Type: application/json" \
+  https://slack.com/api/chat.postMessage \
+  -d '{"channel": "#general", "text": "Hello from reeboot!"}'
+
+# Reply in a thread
+curl -X POST \
+  -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+  -H "Content-Type: application/json" \
+  https://slack.com/api/chat.postMessage \
+  -d '{"channel": "#general", "text": "Reply", "thread_ts": "<parent_ts>"}'
+
+# Get channel history
+curl -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+     "https://slack.com/api/conversations.history?channel=<channel_id>&limit=20"
+
+# Upload a file
+curl -F "file=@/path/to/file.pdf" \
+     -F "channels=<channel_id>" \
+     -H "Authorization: Bearer $SLACK_BOT_TOKEN" \
+     https://slack.com/api/files.upload
+```

package/skills/sqlite/SKILL.md

@@ -0,0 +1,85 @@
+---
+name: sqlite
+description: SQLite database operations via sqlite3 CLI and DATABASE_PATH — run queries, inspect schemas, read and write data. Use when working with a SQLite database file.
+---
+
+# SQLite
+
+Wraps the `sqlite3` CLI for SQLite database operations. Uses the `DATABASE_PATH` environment variable or an explicit file path.
+
+## Setup
+
+1. Install sqlite3 (usually pre-installed on macOS and most Linux distros):
+   ```bash
+   # macOS — check first:
+   sqlite3 --version
+
+   # If not installed:
+   brew install sqlite
+
+   # Ubuntu/Debian:
+   apt-get install sqlite3
+   ```
+
+2. Set the `DATABASE_PATH` environment variable (optional but recommended):
+   ```bash
+   export DATABASE_PATH=/path/to/your/database.db
+   ```
+   Add to your shell profile for persistence. Or pass the path directly in each command.
+
+3. Verify:
+   ```bash
+   sqlite3 "$DATABASE_PATH" ".tables"
+   # or:
+   sqlite3 /path/to/database.db ".tables"
+   ```
+
+## Usage
+
+```bash
+# List all tables
+sqlite3 "$DATABASE_PATH" ".tables"
+
+# Show schema for a table
+sqlite3 "$DATABASE_PATH" ".schema users"
+
+# Show full schema
+sqlite3 "$DATABASE_PATH" ".schema"
+
+# Run a query
+sqlite3 "$DATABASE_PATH" "SELECT * FROM users LIMIT 10;"
+
+# Count rows
+sqlite3 "$DATABASE_PATH" "SELECT COUNT(*) FROM orders WHERE status = 'pending';"
+
+# Insert data
+sqlite3 "$DATABASE_PATH" "INSERT INTO logs (message, created_at) VALUES ('test', datetime('now'));"
+
+# Update data
+sqlite3 "$DATABASE_PATH" "UPDATE users SET active = 1 WHERE email = 'user@example.com';"
+
+# Delete data
+sqlite3 "$DATABASE_PATH" "DELETE FROM sessions WHERE expires_at < datetime('now');"
+
+# Run a SQL file
+sqlite3 "$DATABASE_PATH" < /path/to/migration.sql
+
+# Export to CSV
+sqlite3 -csv "$DATABASE_PATH" "SELECT * FROM users;" > users.csv
+
+# Show database info
+sqlite3 "$DATABASE_PATH" ".dbinfo"
+
+# Check table info (columns, types)
+sqlite3 "$DATABASE_PATH" "PRAGMA table_info(users);"
+
+# Interactive mode
+sqlite3 "$DATABASE_PATH"
+```
+
+### Tips
+
+- Use `LIMIT` on large tables to avoid returning too many rows
+- Use `.headers on` and `.mode column` in interactive mode for readable output
+- SQLite files can be inspected directly with `ls -lh "$DATABASE_PATH"` to check size
+- Use `PRAGMA integrity_check;` to verify database health

package/skills/web-research/SKILL.md

@@ -0,0 +1,47 @@
+---
+name: web-research
+description: Structured multi-query web research using the web-search extension. Use when researching a topic, finding current information, investigating a question, or gathering facts from the web. Run 3–5 targeted searches, synthesise findings, cite URLs.
+---
+
+# Web Research
+
+Structured research pattern using the web-search extension. Produces a synthesis with cited sources.
+
+## Setup
+
+No external dependencies beyond the web-search extension being enabled in your reeboot config.
+
+Verify the extension is active:
+```
+reeboot reload
+```
+
+The `web_search` tool must be available in the agent's tool list. If not, enable the web-search extension in your config:
+```json
+{ "extensions": ["web-search"] }
+```
+
+## Usage
+
+Run 3–5 targeted searches for different facets of the question, then synthesise:
+
+```
+# Step 1: broaden the question into sub-queries
+web_search("topic overview 2025")
+web_search("topic technical details")
+web_search("topic comparison alternatives")
+
+# Step 2: check for recency
+web_search("topic news March 2025")
+
+# Step 3: synthesise
+# Combine findings, note agreement/disagreement across sources, cite URLs.
+```
+
+### Best practices
+
+- Use specific, factual queries rather than open-ended questions
+- Include the year or "latest" to avoid stale results
+- Read the top 3 results per query before moving on
+- Cite every claim with its source URL
+- If results conflict, note the discrepancy and explain which source you trust more

package/templates/models-ollama.json

@@ -0,0 +1,16 @@
+{
+  "providers": [
+    {
+      "id": "ollama",
+      "name": "Ollama (local)",
+      "baseUrl": "http://localhost:11434/v1",
+      "models": [
+        {
+          "id": "{{MODEL_ID}}",
+          "name": "{{MODEL_ID}}",
+          "contextWindow": 8192
+        }
+      ]
+    }
+  ]
+}

package/skills/web-search/SKILL.md

@@ -1,32 +0,0 @@
----
-name: web-search
-description: Search the web using SearXNG. Use when you need to find current information, research a topic, look up facts, or gather information from the internet.
----
-
-# Web Search Skill
-
-Use this skill when you need to search the web for information.
-
-## Instructions
-
-1. Formulate a clear, concise search query for the topic
-2. Use the `web_search` tool with your query
-3. Review the results and extract relevant information
-4. Cite your sources when presenting findings
-
-## When to Use
-
-- Looking up current events or news
-- Researching a technical topic
-- Finding documentation or tutorials
-- Verifying facts or statistics
-- Discovering resources on a subject
-
-## Example
-
-```
-User: What is the current version of Node.js LTS?
-→ web_search("Node.js LTS version 2025")
-→ Review results
-→ Report: "The current Node.js LTS version is X.Y.Z (sources: ...)"
-```