@easbot/agent 0.2.3 → 0.2.5

package/dist/assets/txt/context/prompt/anthropic.txt

@@ -21,7 +21,7 @@ When the user directly asks about easbot (eg. "can easbot do...", "does easbot h
 Prioritize technical accuracy and truthfulness over validating the user's beliefs. Focus on facts and problem-solving, providing direct, objective technical info without any unnecessary superlatives, praise, or emotional validation. It is best for the user if easbot honestly applies the same rigorous standards to all ideas and disagrees when necessary, even if it may not be what the user wants to hear. Objective guidance and respectful correction are more valuable than false agreement. Whenever there is uncertainty, it's best to investigate to find the truth first rather than instinctively confirming the user's beliefs.
 
 # Task Management
-You have access to the TodoWrite tools to help you manage and plan tasks. Use these tools VERY frequently to ensure that you are tracking your tasks and giving the user visibility into your progress.
+You have access to the Todo tools to help you manage and plan tasks. Use these tools VERY frequently to ensure that you are tracking your tasks and giving the user visibility into your progress.
 These tools are also EXTREMELY helpful for planning tasks, and for breaking down larger complex tasks into smaller steps. If you do not use this tool when planning, you may forget to do important tasks - and that is unacceptable.
 
 It is critical that you mark todos as completed as soon as you are done with a task. Do not batch up multiple tasks before marking them as completed.
@@ -30,13 +30,13 @@ Examples:
 
 <example>
 user: Run the build and fix any type errors
-assistant: I'm going to use the TodoWrite tool to write the following items to the todo list:
+assistant: I'm going to use the Todo tool to write the following items to the todo list:
 - Run the build
 - Fix any type errors
 
 I'm now going to run the build using Bash.
 
-Looks like I found 10 type errors. I'm going to use the TodoWrite tool to write 10 items to the todo list.
+Looks like I found 10 type errors. I'm going to use the Todo tool to write 10 items to the todo list.
 
 marking the first todo as in_progress
 
@@ -50,7 +50,7 @@ In the above example, the assistant completes all the tasks, including the 10 er
 
 <example>
 user: Help me write a new feature that allows users to track their usage metrics and export them to various formats
-assistant: I'll help you implement a usage metrics tracking and export feature. Let me first use the TodoWrite tool to plan this task.
+assistant: I'll help you implement a usage metrics tracking and export feature. Let me first use the Todo tool to plan this task.
 Adding the following todos to the todo list:
 1. Research existing metrics tracking in the codebase
 2. Design the metrics collection system
@@ -70,7 +70,7 @@ I've found some existing telemetry code. Let me mark the first todo as in_progre
 # Doing tasks
 The user will primarily request you perform software engineering tasks. This includes solving bugs, adding new functionality, refactoring code, explaining code, and more. For these tasks the following steps are recommended:
 -
-- Use the TodoWrite tool to plan the task if required
+- Use the Todo tool to plan the task if required
 
 - Tool results and user messages may include <system-reminder> tags. <system-reminder> tags contain useful information and reminders. They are automatically added by the system, and bear no direct relation to the specific tool results or user messages in which they appear.
 
@@ -93,7 +93,7 @@ user: What is the codebase structure?
 assistant: [Uses the Task tool]
 </example>
 
-IMPORTANT: Always use the TodoWrite tool to plan and track tasks throughout the conversation.
+IMPORTANT: Always use the Todo tool to plan and track tasks throughout the conversation.
 
 # Code References
 

package/dist/assets/txt/context/template/CODER.txt

@@ -31,12 +31,12 @@ Coder mode is for code development, debugging, and refactoring. The agent must d
 
 ### NEVER
 - Implement features beyond what was requested
-- Use `task` or `plan` tools for subtask decomposition (use `todowrite` instead)
+- Use `task` or `plan` tools for subtask decomposition (use `Todo` instead)
 - Leave code in broken state
 - Skip verification steps
 
 ### ALWAYS
-- Use `todowrite` tool to track multi-step task progress
+- Use `Todo` tool to track multi-step task progress
 - Run lint/typecheck after code changes
 - Create tests for new functionality
 - Verify all tests pass before completion
@@ -94,14 +94,14 @@ Transform tasks into verifiable goals:
 - "Refactor X" → "Ensure tests pass before and after"
 
 **For multi-step tasks, use the Todo tool to track progress:**
-- Use `todowrite` tool to create and update task lists
+- Use `Todo` tool to create and update task lists
 - Update todo status after each step completion
 - Plan steps before implementation
 
 **Example workflow:**
 ```
 1. [Analyze requirements] → verify: [understand user intent]
-2. [Create todo list] → verify: [use todowrite tool to record steps]
+2. [Create todo list] → verify: [use Todo tool to record steps]
 3. [Implement step 1] → verify: [run tests/verify]
 4. [Update todo status] → verify: [confirm progress]
 5. [Implement step 2...] → verify: [run tests/verify]

package/dist/assets/txt/session/prompt/anthropic.txt

@@ -21,7 +21,7 @@ When the user directly asks about easbot (eg. "can easbot do...", "does easbot h
 Prioritize technical accuracy and truthfulness over validating the user's beliefs. Focus on facts and problem-solving, providing direct, objective technical info without any unnecessary superlatives, praise, or emotional validation. It is best for the user if easbot honestly applies the same rigorous standards to all ideas and disagrees when necessary, even if it may not be what the user wants to hear. Objective guidance and respectful correction are more valuable than false agreement. Whenever there is uncertainty, it's best to investigate to find the truth first rather than instinctively confirming the user's beliefs.
 
 # Task Management
-You have access to the TodoWrite tools to help you manage and plan tasks. Use these tools VERY frequently to ensure that you are tracking your tasks and giving the user visibility into your progress.
+You have access to the Todo tools to help you manage and plan tasks. Use these tools VERY frequently to ensure that you are tracking your tasks and giving the user visibility into your progress.
 These tools are also EXTREMELY helpful for planning tasks, and for breaking down larger complex tasks into smaller steps. If you do not use this tool when planning, you may forget to do important tasks - and that is unacceptable.
 
 It is critical that you mark todos as completed as soon as you are done with a task. Do not batch up multiple tasks before marking them as completed.
@@ -30,13 +30,13 @@ Examples:
 
 <example>
 user: Run the build and fix any type errors
-assistant: I'm going to use the TodoWrite tool to write the following items to the todo list:
+assistant: I'm going to use the Todo tool to write the following items to the todo list:
 - Run the build
 - Fix any type errors
 
 I'm now going to run the build using Bash.
 
-Looks like I found 10 type errors. I'm going to use the TodoWrite tool to write 10 items to the todo list.
+Looks like I found 10 type errors. I'm going to use the Todo tool to write 10 items to the todo list.
 
 marking the first todo as in_progress
 
@@ -50,7 +50,7 @@ In the above example, the assistant completes all the tasks, including the 10 er
 
 <example>
 user: Help me write a new feature that allows users to track their usage metrics and export them to various formats
-assistant: I'll help you implement a usage metrics tracking and export feature. Let me first use the TodoWrite tool to plan this task.
+assistant: I'll help you implement a usage metrics tracking and export feature. Let me first use the Todo tool to plan this task.
 Adding the following todos to the todo list:
 1. Research existing metrics tracking in the codebase
 2. Design the metrics collection system
@@ -70,7 +70,7 @@ I've found some existing telemetry code. Let me mark the first todo as in_progre
 # Doing tasks
 The user will primarily request you perform software engineering tasks. This includes solving bugs, adding new functionality, refactoring code, explaining code, and more. For these tasks the following steps are recommended:
 -
-- Use the TodoWrite tool to plan the task if required
+- Use the Todo tool to plan the task if required
 
 - Tool results and user messages may include <system-reminder> tags. <system-reminder> tags contain useful information and reminders. They are automatically added by the system, and bear no direct relation to the specific tool results or user messages in which they appear.
 
@@ -93,7 +93,7 @@ user: What is the codebase structure?
 assistant: [Uses the Task tool]
 </example>
 
-IMPORTANT: Always use the TodoWrite tool to plan and track tasks throughout the conversation.
+IMPORTANT: Always use the Todo tool to plan and track tasks throughout the conversation.
 
 # Code References
 

package/dist/assets/txt/tool/bash.txt

@@ -81,7 +81,7 @@ Git Safety Protocol:
 
 Important notes:
 - NEVER run additional commands to read or explore code, besides git bash commands
-- NEVER use the TodoWrite or Task tools
+- NEVER use the Todo or Task tools
 - DO NOT push to the remote repository unless the user explicitly asks you to do so
 - IMPORTANT: Never use git commands with the -i flag (like git rebase -i or git add -i) since they require interactive input which is not supported.
 - If there are no changes to commit (i.e., no untracked files and no modifications), do not create an empty commit
@@ -108,7 +108,7 @@ gh pr create --title "the pr title" --body "$(cat <<'EOF'
 </example>
 
 Important:
-- DO NOT use the TodoWrite or Task tools
+- DO NOT use the Todo or Task tools
 - Return the PR URL when you're done, so the user can see it
 
 # Other common operations

package/dist/assets/txt/tool/codebase.txt

@@ -1,42 +1,38 @@
 Use this tool to search and analyze code structure in the codebase knowledge graph. It provides code snippets with AST type, language, name, and file location.
 
-**search** — Use this operation to find code by query:
-- Set operation="search"
-- Provide query string to search code content
-- Use astType to filter by code element type (e.g., class_declaration, function_declaration)
-- Use language to filter by programming language (e.g., ts, py, rs)
-- Returns code snippets with related edges when available
-
-**queryEdges** — Use this operation to explore code relationships:
-- Set operation="queryEdges"
-- Provide nodeId to get node information and all connected relations
-- IMPORTANT: nodeId must be in full format: `language:filepath:type:name`
-- Example: `ts:E:/work/apps/eas/project/src/session.ts:class_declaration:Session`
-- Returns node details and incoming/outgoing edges (shown as [in] or [out])
-- Useful for understanding code dependencies and navigation
-
-## Supported Filter Values
-
-**astType** — Code element types:
-`class_declaration`, `class_definition`, `function_declaration`, `function_definition`, `method_definition`, `method_declaration`, `interface_declaration`, `type_alias_declaration`, `enum_declaration`, `variable_declaration`, `import_declaration`, `import_statement`, `struct_declaration`, `struct_specifier`
-
-**language** — Programming languages:
-`ts`, `tsx`, `js`, `jsx`, `py`, `rs`, `go`, `c`, `cpp`, `csharp`, `java`, `scala`, `ruby`, `php`, `zig`
+**search** — Find code by query:
+- operation="search"
+- query: Search query string
+- maxResults: Max results to return (default: 10)
+- astType: Filter by AST type (e.g., class_declaration, function_declaration)
+- language: Filter by language (e.g., ts, py, rs)
+- minScore: Min score threshold (default: 0.3)
+
+**queryEdges** — Explore code relationships:
+- operation="queryEdges"
+- nodeId: Node ID to query relations (required)
+- direction: Edge direction - "in", "out", or "both" (default: "both")
+- maxDepth: Max depth of relations (default: 2)
 
 ## Usage Examples
 
 ```
-# Search for Session class definitions
+# Search for Session class
 codebase(operation="search" query="Session" astType="class_declaration")
 
-# Search for Python functions only
-codebase(operation="search" query="handler" language="py" astType="function_declaration")
+# Search Python functions
+codebase(operation="search" query="handler" language="py")
 
-# Query edges of a specific node (copy nodeId from search results)
-codebase(operation="queryEdges" nodeId="<paste nodeId from search results>")
+# Query node relationships
+codebase(operation="queryEdges" nodeId="ts:/path/file.ts:class:Session")
 ```
 
-## Important Notes
+## Filter Values
+
+**astType** — Code element types:
+`class_declaration`, `class_definition`, `function_declaration`, `function_definition`, `method_definition`, `method_declaration`, `interface_declaration`, `type_alias_declaration`, `enum_declaration`, `variable_declaration`, `import_declaration`, `import_statement`, `struct_declaration`, `struct_specifier`
+
+**language** — Programming languages:
+`ts`, `tsx`, `js`, `jsx`, `py`, `rs`, `go`, `c`, `cpp`, `csharp`, `java`, `scala`, `ruby`, `php`, `zig`
 
-- When searching, the nodeId is returned in results (e.g., `node id=ts:E:/path/to/file.ts:class_declaration:Session`)
-- Use queryEdges to explore relationships between code elements
+**direction**: in (incoming), out (outgoing), both

package/dist/assets/txt/tool/gateway-channel.txt

@@ -1,73 +1,31 @@
-Send messages to external channels through the Gateway.
-
-This tool allows you to proactively send messages to external channels (like Telegram, Discord, Slack) through the Gateway system. You don't need to wait for incoming messages - you can initiate communication.
-
-## Operations
-
-### list
-List all available channels that you can send messages to.
-
-```json
-{
-  "operation": "list"
-}
-```
-
-Returns:
-- `channels`: Array of available channels with their platform, status, and priority
-
-### send
-Send a message to a specific channel or let the system auto-select one.
-
-```json
-{
-  "operation": "send",
-  "content": "Hello from the agent!",
-  "channelId": "telegram-main",  // Optional: specify channel
-  "sessionId": "session-123",    // Optional: for channel selection
-  "messageType": "text"          // "text" or "markdown"
-}
-```
-
-Parameters:
-- `content` (required): The message content to send
-- `channelId` (optional): Specific channel to send to. If not provided:
-  - If `sessionId` is provided, uses the channel bound to that session
-  - Otherwise, auto-selects the highest priority available channel
-- `sessionId` (optional): Session ID for channel binding lookup
-- `messageType` (optional): "text" or "markdown" (default: "text")
-
-## Channel Selection Logic
-
-When `channelId` is not specified:
-1. If `sessionId` is provided and that session is bound to a channel, use that channel
-2. Otherwise, select the highest priority running channel
-
-## Examples
-
-List available channels:
-```
-Use gateway_channel with operation: "list"
-```
-
-Send to a specific channel:
-```
-Use gateway_channel with:
-- operation: "send"
-- channelId: "telegram-main"
-- content: "Hello!"
-```
-
-Send with auto-selection:
-```
-Use gateway_channel with:
-- operation: "send"
-- content: "Hello!"
-```
-
-## Notes
-
-- The Gateway must be running for this tool to work
-- Channels must be registered and running to receive messages
-- Messages are sent proactively - no incoming message is required
-- Use this tool when you need to notify users or send updates to external platforms
+Send messages to external channels through the Gateway system.
+
+**list** — List available channels:
+- operation="list"
+- sessionId: Optional session ID for channel binding lookup
+
+**send** — Send message to channel:
+- operation="send"
+- content: Message content (required)
+- channelId: Specific channel ID (optional, auto-selects if not provided)
+- sessionId: Session ID for channel lookup (optional)
+- messageType: "text", "image", or "file" (default: "text")
+
+## Usage Examples
+
+```
+# List channels
+gateway_channel(operation="list")
+
+# Send message
+gateway_channel(operation="send" content="Hello!")
+
+# Send to specific channel
+gateway_channel(operation="send" channelId="telegram-main" content="Hello!")
+```
+
+## Channel Selection
+
+If channelId not provided:
+1. Use channel bound to sessionId (if provided)
+2. Otherwise auto-select highest priority running channel

package/dist/assets/txt/tool/ls.txt

@@ -1 +1,6 @@
-Lists files and directories in a given path. The path parameter must be absolute; omit it to use the current workspace directory. You can optionally provide an array of glob patterns to ignore with the ignore parameter. You should generally prefer the Glob and Grep tools, if you know which directories to search.
+- Lists all files and directories in a given path
+- Supports tree-style output showing directory hierarchy
+- Automatically ignores common directories like node_modules, .git, dist, build, and more
+- Supports custom ignore patterns via the ignore parameter
+- Returns file count and truncation status when limit is reached
+- When you need to find files by name patterns, prefer the Glob tool

package/dist/assets/txt/tool/note.txt

@@ -1,24 +1,27 @@
 Use this tool to search and retrieve information from the local note/document knowledge base. It returns document paths, line ranges, snippets, and related graph nodes.
 
-**search** — Use this operation to find notes by query:
-- Set operation="search"
-- Provide query string to search note content
-- Set limit to control number of results (default: 10)
-- Returns snippets with document paths and related nodes
+**search** — Find notes by query:
+- operation="search"
+- query: Search query string
+- maxResults: Max results to return (default: 10)
+- minScore: Min score threshold 0-1 (default: 0.3, higher = more precise)
 
-**queryEdges** — Use this operation to explore note relationships:
-- Set operation="queryEdges"
-- Provide nodeId (number) to get node information and all connected neighbors/edges
-- Returns node details and its relationships in the knowledge graph
-- Useful for exploring connections between notes
+**queryEdges** — Explore note relationships:
+- operation="queryEdges"
+- nodeId: Node ID (number) to query relations
+- maxDepth: Max depth of relations (default: 2)
 
-## Supported Parameters
+## Usage Examples
 
-**source** — Note sources (used in config, not directly in tool):
-`docs` — documentation files from note management
-`tasks` — task files from task management
+```
+# Search for notes
+note(operation="search" query="API design" maxResults=10)
 
-Use this tool when you need to:
+# Query node relationships
+note(operation="queryEdges" nodeId=123)
+```
+
+Use when you need to:
 - Find information from project documentation
 - Search for specific topics in notes
 - Explore relationships between documents

package/dist/assets/txt/tool/plan.txt

@@ -0,0 +1,34 @@
+Switch between plan mode and build mode for structured workflow.
+
+Operations:
+- enter: Request to switch to plan mode for research and planning
+- exit: Request to exit plan mode and switch to build mode
+
+**Enter Plan Mode (operation: "enter")**
+
+Call this tool when:
+- The user's request is complex and would benefit from planning first
+- You want to research and design before making changes
+- The task involves multiple files or significant architectural decisions
+- If they explicitly mention wanting to create a plan ALWAYS call this tool first
+
+Do NOT call this tool:
+- For simple, straightforward tasks
+- When the user explicitly wants immediate implementation
+
+After entering plan mode:
+- Research the requirements
+- Create a detailed plan file
+- Use plan_exit tool when planning is complete
+
+**Exit Plan Mode (operation: "exit")**
+
+Call this tool when:
+- You have written a complete plan to the plan file
+- You have clarified any questions with the user
+- You are confident the plan is ready for implementation
+
+Do NOT call this tool:
+- Before you have created or finalized the plan
+- If you still have unanswered questions about the implementation
+- If the user has indicated they want to continue planning

package/dist/assets/txt/tool/pty_session.txt

@@ -1,52 +1,60 @@
-Create and manage persistent interactive PTY sessions for long-running processes. Enables background execution, session persistence, and reconnection capabilities for continuous operations.
-
-**!! CRITICAL WARNING: PERSISTENT BACKGROUND SESSIONS !!**
-
-Key capabilities:
-- Persistent session creation
-- Background process management
-- Session reconnection
-- Multi-session support
-- State preservation
-
-Usage scenarios:
-- Long-running services (web servers, databases)
-- Continuous monitoring (log tailing, system monitoring)
-- Interactive development environments
-- Remote session management
-- Daemon-like processes running persistently
-
-Parameters:
-- command (required): Initial command to run in session
-- sessionName (optional): Unique identifier for session
-- cols/rows (optional): Terminal dimensions
-- autoClose (optional): Auto-close session when idle (default: false)
-
-Session management features:
-- Create persistent sessions that continue running
-- Reconnect to existing sessions
-- Monitor session status and output
-- Manage multiple concurrent sessions
-- Graceful session termination
-
-**!! ABSOLUTELY CRITICAL RESOURCE MANAGEMENT !!**
-
-**WARNING**: Sessions run PERSISTENTLY in background until EXPLICITLY terminated!
-
-**MANDATORY CLEANUP PROCEDURES**:
-1. **MUST use pty_manage** tool for ALL session cleanup operations
-2. **UNUSED sessions CONTINUE consuming system resources indefinitely**
-3. **Sessions do not automatically terminate** - they run indefinitely until explicitly closed
-4. **REGULAR monitoring required**: pty_manage { "action": "list" }
-5. **IMMEDIATE termination**: pty_manage { "action": "close", "sessionId": "id" }
-6. **FAILURE to clean up WILL lead to SYSTEM RESOURCE EXHAUSTION**
-7. **SECURITY RISK**: Uncleaned sessions provide persistent system access
-
-**RESOURCE CONSUMPTION IMPACT**:
-- CPU cycles for active processes
-- Memory allocation for session buffers
-- File descriptors for terminal I/O
-- Network connections (if applicable)
-- System limits may be reached causing crashes
-
-Security: Requires authorization for session creation and management.
+Create and manage persistent interactive PTY sessions for long-running processes. Enables background execution, session persistence, and reconnection capabilities for continuous operations.
+
+**!! CRITICAL WARNING: PERSISTENT BACKGROUND SESSIONS !!**
+
+Key capabilities:
+- Persistent session creation
+- Background process management
+- Session reconnection
+- Multi-session support
+- State preservation
+
+Usage scenarios:
+- Long-running services (web servers, databases)
+- Interactive SSH sessions
+- REPL environments (Node.js, Python, etc.)
+- Continuous monitoring (log tailing, system monitoring)
+- Interactive development environments
+
+Parameters:
+- command (required): Initial command to run in session
+- sessionName (optional): Unique identifier for session
+- cols/rows (optional): Terminal dimensions
+- persistent (optional): Keep session running after command completes (default: true)
+- truncateMode (optional): head or tail for output truncation (default: tail)
+
+**!! INPUT INTERACTION !!**
+
+Send input to session using pty_manage tool:
+```
+pty_manage { "action": "write", "sessionId": "<sessionId>", "data": "your input\n" }
+```
+
+Interactive workflow:
+1. Create session: pty_session { "command": "ssh user@host" }
+2. Send input: pty_manage { "action": "write", "sessionId": "...", "data": "password\n" }
+3. Send commands: pty_manage { "action": "write", "sessionId": "...", "data": "ls -la\n" }
+4. Check output: pty_manage { "action": "get", "sessionId": "..." }
+5. Close session: pty_manage { "action": "close", "sessionId": "..." }
+
+**!! ABSOLUTELY CRITICAL RESOURCE MANAGEMENT !!**
+
+**WARNING**: Sessions run PERSISTENTLY in background until EXPLICITLY terminated!
+
+**MANDATORY CLEANUP PROCEDURES**:
+1. **MUST use pty_manage** tool for ALL session cleanup operations
+2. **UNUSED sessions CONTINUE consuming system resources indefinitely**
+3. **Sessions do not automatically terminate** - they run indefinitely until explicitly closed
+4. **REGULAR monitoring required**: pty_manage { "action": "list" }
+5. **IMMEDIATE termination**: pty_manage { "action": "close", "sessionId": "id" }
+6. **FAILURE to clean up WILL lead to SYSTEM RESOURCE EXHAUSTION**
+7. **SECURITY RISK**: Uncleaned sessions provide persistent system access
+
+**RESOURCE CONSUMPTION IMPACT**:
+- CPU cycles for active processes
+- Memory allocation for session buffers
+- File descriptors for terminal I/O
+- Network connections (if applicable)
+- System limits may be reached causing crashes
+
+Security: Requires authorization for session creation and management.