- Introduce `conn sso` CLI suite for managing Identity Providers (IdP).
- Implement `login_sso` and `get_sso_providers` in gRPC AuthService.
- Add auto-provisioning for users logging in via SSO.
- Support JWT validation via shared secrets (HS256) or JWKS (RS256).
- Add domain restriction (`allowed_domains`) and env-var secret resolution.
- Increase JWT session expiration from 8 to 12 hours.
- Add shell autocompletion for SSO commands and configured providers.
- Bump version to 6.0.3.
- Implement dynamic regex matching fallback (re.search) in `node.test` with safe handling of invalid patterns.
- Refactor terminal window resizing (setwinsize) to trigger only on non-router devices and handle SIGWINCH re-renders.
- Introduce `contextvars` context copying for background worker threads in gRPC execution and AI servicers.
- Add unit tests for regex validation, malformed expression fallbacks, and variable formatting in node testing.
- Optimize Playbook Builder AI guidelines for single-task test evaluations.
- Unify codebase comments to English.
- Priority Matching: Prioritize exact node matches in connect, delete, show, and modify actions to bypass disambiguation prompts and prevent accidental bulk mutations on partial matches.
- Bulk Operations: Optimize NodeService delete and update operations by deferring configuration writes, syncs, and cache updates to the final element of a batch.
- Remote AI: Prevent client-side CLI deadlocks when the gRPC server encounters AI configuration ValueErrors by returning a clean error state and final stream marker.
- Testing: Add unit test to verify exact-match priority behavior and update existing CLI tests to match new NodeService signatures.
- Servidor gRPC: Agregar interceptores de autenticación y UserRegistry para aislar sesiones por usuario.
- Contexto de Hilos: Corregir propagación de ContextVar _current_user a hilos secundarios en ExecutionServicer.
- Configuración Compartida: Implementar herencia y deep merge de settings de IA ('ai') y servidores MCP en configfile.
- Hot-Reload: Recarga automática en caliente de la configuración compartida global ante cambios en disco.
- CLI: Agregar comandos e interfaces de usuario para autenticación (login) y administración de usuarios.
- Pruebas: Desarrollar tests unitarios completos (test_shared_ai.py) y resolver regresiones en la suite existente.
- AI & Session Management:
- Add random suffix to session IDs to ensure uniqueness.
- Implement optional pagination/limit in session listing (default 20).
- Add `--all` flag to `ai` CLI commands to view all sessions.
- Keep active session ID and path synced correctly during clean session startups.
- CLI UI/UX:
- Add a custom `ConnpyTheme` for inquirer prompts that dynamically translates
active hex style colors to terminal ANSI/blessed escapes.
- gRPC & Services:
- Implement remote MCP server listing (`list_mcp_servers` RPC and services).
- Stream responder updates (`__RESPONDER__`) to toggle headers dynamically between
"Network Engineer" and "Network Architect" on remote/web clients.
- Fix remote client deadlock risk by ensuring `final_mark` is sent on exceptions.
- Hydrate client-side chat history correctly on initial streaming request.
- Testing:
- Add integration tests for AI gRPC services and MCP server listing.
Focus: UI stability and efficient rendering.
* Markdown Rendering Rewrite: Removed the dependency on rich.live.Live (which caused flickering and
high CPU usage by constantly re-rendering the entire panel).
* New BlockMarkdownRenderer: Implemented in printer.py (alias IncrementalMarkdownParser), it
accumulates text in a buffer and only prints to the screen when it detects a complete block (e.g.,
line breaks or
` code blocks).
* UI Optimizations (terminal_ui.py & stubs.py): Waiting spinners now stop cleanly, and the UI
transitions smoothly to block printing. Fixed visual truncation issues in the bottom "Tab prompt"
bar for excessively long commands.
📦 2. Commit History (Last 4 Commits)
9446baf - improve ai rules
* Strict Anti-Hallucination (ai.py): Injected MANDATORY rules into the System Prompt for both
Architect and Engineer agents. Now, if the terminal buffer is empty or only contains an idle prompt
(e.g., iol#), the AI is strictly forced to state that it lacks data instead of inventing topologies
or configurations.
* Language Preference: Explicitly instructed agents to always respond in the same language the user
used to ask the question.
64377f7 - move context block logic to server and improvements
* Context Precision (ai_service.py): Moved context partitioning logic to the service/server side. It
now calculates exact start_pos and end_pos based on identified commands, preventing mixed outputs
or residual text from bleeding into the AI's prompt.
* Token Savings (server.py): The server now selectively strips garbage metadata and UI caches that
add no value to the LLM before sending the payload over the wire.
e4fd1ad - fix logclean for 6wind
* Full ANSI/CSI Support (utils.py): Replaced the legacy rigid escape filter with a complete CSI
(Control Sequence Introducer) parser. The client can now accurately process numeric cursor
movements (C, D), inline dynamic erasures (K), and absolute shifts (G), ensuring connpy understands
exactly what a 6WIND router or other VNFs render on the screen without garbage characters.
b0a914a - updates al copilot
* Copilot <> gRPC Integration: Implemented the async plumbing required for the Copilot to work over
the gRPC tunnel (server.py & stubs.py).
* Initial Streaming UI: Added the first iteration of chunk callbacks to provide real-time feedback
while the LLM generates responses, laying the groundwork for the uncommitted block-renderer
optimizations.
pointing to /config. This avoids running the conn command during the build process and ensures a
cleaner setup.
2. Copilot UI Fix: Resolved a double-escaping bug in the terminal bottom bar. Device prompts (like
6WIND-PE1>) will now render correctly instead of showing HTML entities like >.
3. AI Model Update: Updated the default engineer model in connpy/ai.py to
gemini/gemini-3.1-flash-lite, removing the deprecated -preview suffix.
4. Standardized Timeouts: Unified all default timeouts to 20 seconds across the board. This includes
direct execution (run/test), modern playbooks (v2), and classic task-based playbooks (v1).
5. Documentation Update: Regenerated the full documentation site in the docs/ directory using pdoc to
reflect the latest changes.
6. Cleanup: Removed all debug prints from connpy/core.py and restored the docker/logs/.gitignore
file.
- AI Concurrency:
- Implemented a dedicated background event loop (ConnpyAILoop) in a separate thread for AI tasks to ensure thread safety and event loop affinity.
- Added 'run_ai_async' utility to funnel all LiteLLM calls through the dedicated loop.
- Implemented global 'cleanup()' for safe closure of sync/async LiteLLM sessions.
- gRPC & Remote Sessions:
- Enhanced 'NodeServicer' to identify command blocks within the terminal buffer using prompt regex/byte tracking.
- Added support for selective context retrieval via 'context_start_pos' in the gRPC Interact stream.
- Synchronized remote Copilot behavior by enriching questions with session history (last 5 queries) in 'NodeStub'.
- Optimized token usage by cleaning 'node_info' metadata before AI transmission.
- Terminal Context & Core:
- Modified 'node.connect' to always initialize 'mylog' (BytesIO) buffer regardless of disk logging configuration, ensuring Copilot context availability.
- Integrated 'ai.cleanup()' in CLI (connapp) and Server (api) exit points for graceful shutdowns.
- Suppressed LiteLLM internal streaming coroutine warnings during task cancellation.
- Implement threaded plugin execution with Queue-based streaming in PluginService
- Refactor remote logger to preserve ANSI colors and fix TTY line endings (\r\n)
- Intelligent terminal filtering: disable SSM screen-clearing filter after success
- Sanitize SSH-only flags in core.py when using SFTP protocol
- Rewrite completion tree with pre/post-node states and flag deduplication
- Update gRPC unit tests to match new streaming response structure
- **Core/Jumphost**: Implement native jumphost support for `ssm`, `kubectl`, and `docker` protocols via transparent ProxyCommands.
- **Core/Tags**: Introduce `ssh_options` tag to pass custom arguments (e.g., `-i key.pem`) inside nested SSM SSH tunnels.
- **Core/Tags**: Introduce `nc_command` tag to override default `nc` behavior in Docker/Kube tunnels (crucial for IOS-XRd VRF netns isolation).
- **Core/Async**: Fix a critical bug where interactive terminal sessions disconnected immediately due to premature `asyncio.FIRST_COMPLETED` task resolution when logs or idletime were disabled.
- **Services/Node**: Fix logic error in `validate_parent_folder` that prevented the creation of new nested subfolders (CLI was incorrectly checking for the subfolder's own existence instead of the parent's).
- **CLI/Help**: Update documentation to include the new `region`, `profile`, `ssh_options`, and `nc_command` well-known tags.
- Replaced blocking pexpect interact with asyncio-based multiplexing.
- Implemented LocalStream and RemoteStream for agnostic I/O handling.
- Real-time SIGWINCH window resizing support.
- 'Ghost buffer' mitigation for clean, artifact-free session handovers.
- Upgraded _logclean into a mini terminal emulator to accurately process ANSI, backspaces, and inline clears.
- Continuous auto-saving for logs without blocking the main thread.
- Bumped version to 6.0.0b1 and regenerated pdoc documentation.
- Bump version to 5.0b6 and regenerate HTML documentation via pdoc3.
- Add persistent AI chat sessions (list, resume, delete) stored locally.
- Fix 'rich' library console rendering and routing 'error()' to stderr.
- Update Architect UI color theme to medium_purple.
- Sanitize caching metadata (cache_control) for compatibility with non-Anthropic models.
- Fix .folder config path redirection mapping and fzf-wrapper compatibility.
- Ensure context plugin correctly filters node lists upon load.
- Inject config instance directly into API components instead of instantiating globally.
- Fix edge-case in plugin loading preventing startup when folder is missing.
- Add comprehensive test coverage for printer module and AI sessions.
- Refactored completion.py to use text caches for near-instant tab-completion.
- Integrated self-healing cache generation in configfile.py for nodes, folders, and profiles.
- Updated bash/zsh completion generation to call completion.py directly via python3.
- Embedded Google OAuth client config in sync.py with split secrets to bypass GitHub scanning.
- Refactored context plugin with property-based configuration and improved safety (default 'all' context, regex fallback).
- Updated unit tests for completion caching and validated context plugin improvements.
- Bumped version to 5.0b4 and regenerated documentation.
- Migrated configuration backend from JSON to YAML for better readability.
- Added automatic dual-caching (.config.cache.json) to preserve fast load times with YAML.
- Implemented a new 0ms latency fzf wrapper for bash and zsh (--fzf-wrapper).
- Updated sync plugin to support the new YAML config format and clear caches on extraction.
- Refactored 'completion.py' to gracefully handle fallback config formats.
- Added new test modules (test_capture, test_context, test_sync) covering core plugins.
- Updated existing unit tests to handle YAML config creation and parsing.
- Bumped version to 5.0b3 and regenerated HTML documentation.
- Added comprehensive unit test suite covering AI, core, plugins, completion, API, and hooks using pytest.
- Replaced bare 'except:' clauses across the codebase with specific exception handling (e.g., ValueError, KeyError, OSError) to prevent swallowing system exit calls.
- Fixed Python 3.8+ AST compatibility in plugins.py (support for ast.Constant).
- Removed deprecated pkg_resources import from __init__.py.
- Fixed missing 'printer' import in configfile.py that caused NameErrors during save failures.
- Refactored AI module to use litellm, supporting Anthropic, Google, OpenAI, etc.
- Introduced 'Engineer' (execution) and 'Architect' (strategic) AI agents.
- Added real-time streaming responses and interactive chat mode via 'rich'.
- Added CLI arguments for model/key overrides (--engineer-model, --architect-model).
- Replaced 'openai' with 'litellm' in requirements.txt and setup.cfg.
- Updated nodes.run() to support an 'on_complete' callback for live node-status streaming.
- Fixed an undefined variable bug (config.profiles -> self.profiles) in configfile.py.
- Updated README and docstrings with new AI plugin tool registration API.
- Regenerated HTML documentation using pdoc3.
- Bumped version to 5.0b1 for beta release.
fluzzi32
