omnimancer-cli 0.2.22__tar.gz → 0.2.24__tar.gz

{omnimancer_cli-0.2.22/omnimancer_cli.egg-info → omnimancer_cli-0.2.24}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: omnimancer-cli
-Version: 0.2.22
+Version: 0.2.24
 Summary: A unified command-line interface for multiple AI language models
 Author-email: Kellan Strong <kellan@jite.ai>
 License: MIT
@@ -37,7 +37,8 @@ Requires-Dist: jsonschema>=4.17.0
 Requires-Dist: readability-lxml>=0.8.1
 Requires-Dist: PyYAML>=6.0.0
 Requires-Dist: toml>=0.10.2
-Requires-Dist: mcp>=1.0.0
+Requires-Dist: mcp<2,>=1.0.0
+Requires-Dist: prompt-toolkit<4.0,>=3.0.43
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0.0; extra == "dev"
 Requires-Dist: pytest-cov>=4.0.0; extra == "dev"
@@ -49,9 +50,11 @@ Requires-Dist: flake8>=6.0.0; extra == "dev"
 Requires-Dist: mypy>=1.0.0; extra == "dev"
 Dynamic: license-file
 
-# Omnimancer
+<p align="center">
+  <img src="assets/omnimancer.png" alt="Omnimancer" width="600">
+</p>
 
-A multi-model coding agent for the terminal. One tool, any LLM.
+<p align="center"><strong>A multi-model coding agent for the terminal. One tool, any LLM.</strong></p>
 
 Omnimancer works like `claude -p` but isn't locked to a single provider. Point it at Claude, OpenAI, Gemini, Bedrock, Ollama, or any of 13+ supported backends and get a coding agent that reads files, writes code, runs commands, and iterates autonomously — with streaming responses, token/cost tracking, and structured JSON output for pipeline integration.
 
@@ -62,6 +65,10 @@ Beyond the basics, it ships with:
 - **Permission rules** — declarative allow/deny/ask rules per tool with regex matchers
 - **Subagents** — scoped child agents with their own prompt, tool whitelist, and model
 - **Layered security** — approval workflow, sensitive-path protection, and project-boundary enforcement
+- **Named endpoint aliases** — run several OpenAI-compatible endpoints (vLLM, llama.cpp, LM Studio) side by side as first-class providers, keyless if the endpoint is
+- **Modern input layer** — prompt_toolkit prompt with multiline editing, fuzzy `@file` mentions with content injection, live completion for providers/models, Ctrl-R history search
+- **Prompt enhancement** — `e:` prefix or `/enhance` rewrites your draft with a configurable small model before sending (PromptFoundry meta-prompts)
+- **Session approval modes** — `/accept edits|all` (or Shift+Tab) for claude-code-style auto-accept, always below your permission rules
 
 ## Install
 
@@ -166,6 +173,7 @@ Streaming is automatic for providers that support it. Other providers fall back
 | Provider | Streaming |
 |----------|:---------:|
 | Claude (Anthropic) | Yes |
+| OpenAI / openai-compatible / DigitalOcean | Yes (SSE, incl. tool-call deltas) |
 | All others | Fallback (full response) |
 
 Streaming works in both regular chat and agent mode (tool calling flow). The display uses a live-updating terminal panel that refreshes at 15fps.
@@ -244,7 +252,8 @@ Tool names match the agent's toolset (`Read`, `Write`, `Edit`, `Bash`, `Glob`, `
 | `/quit` | Exit (also: `/exit`, Ctrl+D) |
 | `/clear` | Clear terminal screen |
 | `/switch <provider> [model]` | Switch provider or model (uncataloged models are registered on the fly) |
-| `/models [filter]` | List available models (alias: `/model`) |
+| `/models [filter]` | List available models; `/models refresh [provider]` pulls live catalogs |
+| `/model` | Interactive model picker for the current provider |
 | `/providers` | List all providers with status |
 | `/agent on\|off\|status` | Toggle agent mode |
 | `/subagents [run <name> <task>]` | List or run scoped child agents |
@@ -253,6 +262,8 @@ Tool names match the agent's toolset (`Read`, `Write`, `Edit`, `Bash`, `Glob`, `
 | `/config remove-provider <name>` | Remove a provider |
 | `/hooks [list\|on\|off\|add\|remove]` | Manage lifecycle hooks |
 | `/permissions [list\|on\|off\|allow\|deny\|ask\|remove]` | Manage permission rules |
+| `/accept [edits\|all\|off]` | Session approval mode (Shift+Tab cycles it at the prompt) |
+| `/enhance [chat\|code\|image\|research] <draft>` | Rewrite a draft prompt with the enhancement model, confirm, send (the `e:` message prefix does it inline) |
 | `/save [name]` | Save conversation |
 | `/load [name]` | Load conversation |
 | `/list` | List saved conversations |
@@ -426,6 +437,39 @@ several endpoints side by side and switch between them with `/switch <provider>`
 Any OpenAI-compatible service (local proxy, gateway, self-hosted model) works by
 overriding `base_url` on the `openai` provider.
 
+
+### Self-hosted endpoints as named providers
+
+Any number of OpenAI-compatible endpoints (vLLM, llama.cpp server, LM Studio, proxies) can be configured side by side under names you choose. `provider_type` picks the implementation; the entry name is yours:
+
+```bash
+>>> /config set-provider gateway --type openai-compatible \
+        --base-url http://vllm-gateway.internal:8888/v1 --model qwen3-coder-30b
+>>> /config set-provider local --type openai-compatible \
+        --base-url http://localhost:8000/v1 --model qwen3-coder-30b
+>>> /switch local
+```
+
+The `openai-compatible` type is keyless by default (`auth_type: "none"`), takes its model catalog from the endpoint's `/v1/models` (including each model's served context size — see `/models refresh`), and raises timeout errors with a cold-start hint since self-hosted gateways may load a model on first request. Set `providers.<name>.timeout` generously for such endpoints (e.g. `360`).
+
+### Prompt enhancement (optional)
+
+`/enhance` and the `e:` message prefix rewrite a draft prompt with a dedicated model before sending, using the [PromptFoundry](https://github.com/gitshipdone/promptfoundry) meta-prompts (same author; that repo is the canonical source for the prompt texts and is being prepared for public release).
+
+The feature is **opt-in**: it activates only when your config carries an `enhancement` block. Without one, `e:` is not intercepted and `/enhance` points you here.
+
+```json
+"enhancement": {
+  "provider": "gateway",
+  "model": "qwen3-8b",
+  "temperature": 0.4,
+  "default_profile": "code",
+  "enabled": true
+}
+```
+
+`provider` must name an entry in `providers`; `default_profile` is one of `chat`, `code`, `image`, `research`. Failsafe chain: the configured model is tried first; if it is unreachable, the model you are currently chatting with rewrites the draft instead; if everything fails, the original draft is sent unchanged — enhancement never blocks a message.
+
 ### Environment variable overrides
 
 Environment variables take precedence over the saved config and are applied at
@@ -571,3 +615,7 @@ Tests follow TDD. 1,460 tests across providers, CLI, streaming, agent operations
 ## License
 
 MIT
+
+---
+
+*This documentation was reviewed by a local Qwen model as part of OMN testing harness.*

{omnimancer_cli-0.2.22 → omnimancer_cli-0.2.24}/README.md

@@ -1,6 +1,8 @@
-# Omnimancer
+<p align="center">
+  <img src="assets/omnimancer.png" alt="Omnimancer" width="600">
+</p>
 
-A multi-model coding agent for the terminal. One tool, any LLM.
+<p align="center"><strong>A multi-model coding agent for the terminal. One tool, any LLM.</strong></p>
 
 Omnimancer works like `claude -p` but isn't locked to a single provider. Point it at Claude, OpenAI, Gemini, Bedrock, Ollama, or any of 13+ supported backends and get a coding agent that reads files, writes code, runs commands, and iterates autonomously — with streaming responses, token/cost tracking, and structured JSON output for pipeline integration.
 
@@ -11,6 +13,10 @@ Beyond the basics, it ships with:
 - **Permission rules** — declarative allow/deny/ask rules per tool with regex matchers
 - **Subagents** — scoped child agents with their own prompt, tool whitelist, and model
 - **Layered security** — approval workflow, sensitive-path protection, and project-boundary enforcement
+- **Named endpoint aliases** — run several OpenAI-compatible endpoints (vLLM, llama.cpp, LM Studio) side by side as first-class providers, keyless if the endpoint is
+- **Modern input layer** — prompt_toolkit prompt with multiline editing, fuzzy `@file` mentions with content injection, live completion for providers/models, Ctrl-R history search
+- **Prompt enhancement** — `e:` prefix or `/enhance` rewrites your draft with a configurable small model before sending (PromptFoundry meta-prompts)
+- **Session approval modes** — `/accept edits|all` (or Shift+Tab) for claude-code-style auto-accept, always below your permission rules
 
 ## Install
 
@@ -115,6 +121,7 @@ Streaming is automatic for providers that support it. Other providers fall back
 | Provider | Streaming |
 |----------|:---------:|
 | Claude (Anthropic) | Yes |
+| OpenAI / openai-compatible / DigitalOcean | Yes (SSE, incl. tool-call deltas) |
 | All others | Fallback (full response) |
 
 Streaming works in both regular chat and agent mode (tool calling flow). The display uses a live-updating terminal panel that refreshes at 15fps.
@@ -193,7 +200,8 @@ Tool names match the agent's toolset (`Read`, `Write`, `Edit`, `Bash`, `Glob`, `
 | `/quit` | Exit (also: `/exit`, Ctrl+D) |
 | `/clear` | Clear terminal screen |
 | `/switch <provider> [model]` | Switch provider or model (uncataloged models are registered on the fly) |
-| `/models [filter]` | List available models (alias: `/model`) |
+| `/models [filter]` | List available models; `/models refresh [provider]` pulls live catalogs |
+| `/model` | Interactive model picker for the current provider |
 | `/providers` | List all providers with status |
 | `/agent on\|off\|status` | Toggle agent mode |
 | `/subagents [run <name> <task>]` | List or run scoped child agents |
@@ -202,6 +210,8 @@ Tool names match the agent's toolset (`Read`, `Write`, `Edit`, `Bash`, `Glob`, `
 | `/config remove-provider <name>` | Remove a provider |
 | `/hooks [list\|on\|off\|add\|remove]` | Manage lifecycle hooks |
 | `/permissions [list\|on\|off\|allow\|deny\|ask\|remove]` | Manage permission rules |
+| `/accept [edits\|all\|off]` | Session approval mode (Shift+Tab cycles it at the prompt) |
+| `/enhance [chat\|code\|image\|research] <draft>` | Rewrite a draft prompt with the enhancement model, confirm, send (the `e:` message prefix does it inline) |
 | `/save [name]` | Save conversation |
 | `/load [name]` | Load conversation |
 | `/list` | List saved conversations |
@@ -375,6 +385,39 @@ several endpoints side by side and switch between them with `/switch <provider>`
 Any OpenAI-compatible service (local proxy, gateway, self-hosted model) works by
 overriding `base_url` on the `openai` provider.
 
+
+### Self-hosted endpoints as named providers
+
+Any number of OpenAI-compatible endpoints (vLLM, llama.cpp server, LM Studio, proxies) can be configured side by side under names you choose. `provider_type` picks the implementation; the entry name is yours:
+
+```bash
+>>> /config set-provider gateway --type openai-compatible \
+        --base-url http://vllm-gateway.internal:8888/v1 --model qwen3-coder-30b
+>>> /config set-provider local --type openai-compatible \
+        --base-url http://localhost:8000/v1 --model qwen3-coder-30b
+>>> /switch local
+```
+
+The `openai-compatible` type is keyless by default (`auth_type: "none"`), takes its model catalog from the endpoint's `/v1/models` (including each model's served context size — see `/models refresh`), and raises timeout errors with a cold-start hint since self-hosted gateways may load a model on first request. Set `providers.<name>.timeout` generously for such endpoints (e.g. `360`).
+
+### Prompt enhancement (optional)
+
+`/enhance` and the `e:` message prefix rewrite a draft prompt with a dedicated model before sending, using the [PromptFoundry](https://github.com/gitshipdone/promptfoundry) meta-prompts (same author; that repo is the canonical source for the prompt texts and is being prepared for public release).
+
+The feature is **opt-in**: it activates only when your config carries an `enhancement` block. Without one, `e:` is not intercepted and `/enhance` points you here.
+
+```json
+"enhancement": {
+  "provider": "gateway",
+  "model": "qwen3-8b",
+  "temperature": 0.4,
+  "default_profile": "code",
+  "enabled": true
+}
+```
+
+`provider` must name an entry in `providers`; `default_profile` is one of `chat`, `code`, `image`, `research`. Failsafe chain: the configured model is tried first; if it is unreachable, the model you are currently chatting with rewrites the draft instead; if everything fails, the original draft is sent unchanged — enhancement never blocks a message.
+
 ### Environment variable overrides
 
 Environment variables take precedence over the saved config and are applied at
@@ -520,3 +563,7 @@ Tests follow TDD. 1,460 tests across providers, CLI, streaming, agent operations
 ## License
 
 MIT
+
+---
+
+*This documentation was reviewed by a local Qwen model as part of OMN testing harness.*

{omnimancer_cli-0.2.22 → omnimancer_cli-0.2.24}/omnimancer/__init__.py

@@ -10,10 +10,18 @@ try:
 
     __version__ = _get_version("omnimancer-cli")
 except Exception:
-    with open("pyproject.toml", "r") as f:
-        project = f.readlines()
-    version = [i for i in project if "version" in i].pop(0).strip()
-    __version__ = version.split().pop().strip("\"'")
+    # Uninstalled checkout (PYTHONPATH usage): read pyproject.toml next to
+    # the package, never relative to cwd — the CLI can be invoked from any
+    # working directory.
+    try:
+        from pathlib import Path
+
+        _pyproject = Path(__file__).resolve().parent.parent / "pyproject.toml"
+        project = _pyproject.read_text().splitlines()
+        version = [i for i in project if "version" in i].pop(0).strip()
+        __version__ = version.split().pop().strip("\"'")
+    except Exception:
+        __version__ = "unknown"
 __author__ = "Omnimancer Team"
 __description__ = "Unified CLI for multiple AI language models"
 

{omnimancer_cli-0.2.22 → omnimancer_cli-0.2.24}/omnimancer/cli/approval_integration.py

@@ -8,6 +8,7 @@ in terminal with clear formatting and captures approval decisions.
 
 import logging
 from datetime import datetime
+from enum import Enum
 from pathlib import Path
 from typing import Any, Dict, List, Optional
 
@@ -27,6 +28,32 @@ from .approval_prompt import ApprovalDecision, CLIApprovalPrompt
 logger = logging.getLogger(__name__)
 
 
+class ApprovalMode(Enum):
+    """Session-wide approval mode (claude-code-style accept toggles).
+
+    Sits BELOW permission rules: DENY rules refuse before the approval
+    workflow is reached, and ASK rules set ``_force_prompt``, which is
+    checked before the mode. The low-level security gate re-validates
+    paths/commands at execute time regardless.
+    """
+
+    NORMAL = "normal"
+    ACCEPT_EDITS = "accept-edits"
+    ACCEPT_ALL = "accept-all"
+
+
+#: /accept cycle order.
+_MODE_CYCLE = [
+    ApprovalMode.NORMAL,
+    ApprovalMode.ACCEPT_EDITS,
+    ApprovalMode.ACCEPT_ALL,
+]
+
+#: Operation types auto-approved by ACCEPT_EDITS. Deliberately excludes
+#: deletes and command execution — those stay interactive.
+_EDIT_OPERATION_TYPES = frozenset({OperationType.FILE_WRITE})
+
+
 class CLIApprovalIntegration:
     """
     Integration layer that connects EnhancedApprovalManager with CLI approval dialogs.
@@ -60,6 +87,7 @@ class CLIApprovalIntegration:
         self.console = console or Console()
         self.enable_auto_approval = enable_auto_approval
         self.approval_timeout_seconds = approval_timeout_seconds
+        self.session_approval_mode = ApprovalMode.NORMAL
 
         # Initialize UI components
         self.formatter = CLIApprovalFormatter(self.console)
@@ -337,6 +365,18 @@ class CLIApprovalIntegration:
         if operation.data.get("_force_prompt"):
             return None
 
+        # Session approval mode (/accept): accept-edits auto-approves file
+        # writes only; accept-all approves everything that reaches the
+        # approval workflow (deny rules and hard security limits never do).
+        if self._mode_auto_approves(operation):
+            self.console.print(
+                "[dim]✅ Auto-approved:"
+                f" {operation.description}"
+                f" (session mode: {self.session_approval_mode.value})"
+                "[/dim]"
+            )
+            return True
+
         try:
             # Generate operation signature for matching
             operation_signature = self._generate_operation_signature(operation)
@@ -366,6 +406,20 @@ class CLIApprovalIntegration:
             logger.error(f"Error checking auto-approval: {e}")
             return None
 
+    def _mode_auto_approves(self, operation: Operation) -> bool:
+        """Whether the session approval mode auto-approves this operation."""
+        if self.session_approval_mode is ApprovalMode.ACCEPT_ALL:
+            return True
+        if self.session_approval_mode is ApprovalMode.ACCEPT_EDITS:
+            return operation.type in _EDIT_OPERATION_TYPES
+        return False
+
+    def cycle_approval_mode(self) -> ApprovalMode:
+        """Advance normal → accept-edits → accept-all → normal."""
+        index = _MODE_CYCLE.index(self.session_approval_mode)
+        self.session_approval_mode = _MODE_CYCLE[(index + 1) % len(_MODE_CYCLE)]
+        return self.session_approval_mode
+
     async def _store_approval_pattern(
         self,
         operation: Operation,