How huzaifaashraf13/penzer-cli works
Penzer CLI is an interactive shell agent: a user submits a task, the CLI runs the agent loop, the agent consults the model and memory, and tool execution produces results for the terminal. Its major branches include model configuration, reusable skills and plugins, persistent memory, browser automation, and terminal/file operations. The graph follows relationships named by the SOURCE INDEX and README; unsampled subsystem links are omitted where their callers cannot be verified.
Main entry points: agent.py Agent orchestration · execution.py Tool dispatch · cli.py Interactive CLI · tool.py Browser tool
Components 23
CLI and configuration
- Interactive CLI
cli.py - Settings and credentials
config.py - Terminal interface
ui/terminal.py
Agent reasoning
- Agent orchestration
agent/agent.py - LLM interface
agent/llm.py - System prompts
agent/system_prompts.py - Tool dispatch
agent/penzermodule/execution.py - Skill loading and runtime
agent/skills/loader.py
Persistent memory
- Memory orchestration
session/memory.py - Memory semantics
session/memory_core.py - Memory graph
session/memory_graph.py - Memory storage
session/memory_storage.py
Tools and execution
- Plugin discovery
tools/plugins/__init__.py - Safe command executor
tools/executor.py - Terminal tool
tools/terminal/tool.py - File editor
tools/file_editor/tool.py
Browser automation
- Browser tool
tools/browser/tool.py - Browser runtime
tools/browser/runtime.py - Web search
tools/browser/search.py
Other components
- Shell user
- Local or cloud model
- Chrome browser
- Web search providers
Connections 26
- Shell user → Interactive CLI: submits tasks
- Interactive CLI → Terminal interface: renders interface
cli.py - Interactive CLI → Settings and credentials: loads settings
cli.py - Interactive CLI → Agent orchestration: starts agent
cli.py - Interactive CLI → Memory orchestration: accesses memory
cli.py - Agent orchestration → LLM interface: requests completions
agent/agent.py - Agent orchestration → System prompts: builds prompts
agent/agent.py - Agent orchestration → Tool dispatch: dispatches tools
agent/agent.py - Agent orchestration → Skill loading and runtime: uses skills
agent/agent.py - Agent orchestration → Plugin discovery: loads plugins
agent/agent.py - Agent orchestration → Memory orchestration: reads and writes
agent/agent.py - LLM interface → Local or cloud model: calls model API
- Tool dispatch → Safe command executor: runs commands
agent/penzermodule/execution.py - Tool dispatch → Plugin discovery: dispatches plugins
agent/penzermodule/execution.py - Tool dispatch → Terminal tool: dispatches terminal
agent/penzermodule/execution.py - Tool dispatch → File editor: dispatches edits
agent/penzermodule/execution.py - Tool dispatch → Browser tool: dispatches browser
agent/penzermodule/execution.py - Memory orchestration → Memory semantics: uses semantics
session/memory.py - Memory orchestration → Memory graph: uses graph
session/memory.py - Memory semantics → Memory storage: reads and writes
session/memory_core.py - Memory graph → Memory storage: reads and writes
session/memory_graph.py - Terminal tool → Safe command executor: executes commands
tools/terminal/tool.py - Browser tool → Browser runtime: uses runtime
tools/browser/tool.py - Browser tool → Web search: searches web
tools/browser/tool.py - Browser tool → Chrome browser: controls browser
- Web search → Web search providers: queries providers
tools/browser/search.py