Merge pull request #49 from LittleCoinCoin/dev

Mistral Vibe MCP config support

Author: LittleCoinCoin

README.md

@@ -28,6 +28,7 @@ Hatch supports deployment to the following MCP host platforms:
 - **Codex** — OpenAI Codex with MCP server configuration support
 - **LM Studio** — Local LLM inference platform with MCP server integration
 - **Google Gemini CLI** — Command-line interface for Google's Gemini model with MCP support
+- **Mistral Vibe** — Mistral Vibe CLI coding agent
 
 ## Quick Start
 

__reports__/mistral_vibe/00-parameter_analysis_v0.md

@@ -0,0 +1,55 @@
+# Mistral Vibe Parameter Analysis
+
+## Model
+
+| Item | Finding |
+| --- | --- |
+| Host | Mistral Vibe |
+| Config path | `./.vibe/config.toml` first, fallback `~/.vibe/config.toml` |
+| Config key | `mcp_servers` |
+| Structure | TOML array-of-tables: `[[mcp_servers]]` |
+| Server identity | Inline `name` field per entry |
+
+## Field Summary
+
+| Category | Fields |
+| --- | --- |
+| Transport | `transport`, `command`, `args`, `url` |
+| Common | `headers`, `prompt`, `startup_timeout_sec`, `tool_timeout_sec`, `sampling_enabled` |
+| Auth | `api_key_env`, `api_key_header`, `api_key_format` |
+| Local-only | `env` |
+
+## Host Spec
+
+```yaml
+host: mistral-vibe
+format: toml
+config_key: mcp_servers
+config_paths:
+  - ./.vibe/config.toml
+  - ~/.vibe/config.toml
+transport_discriminator: transport
+supported_transports:
+  - stdio
+  - http
+  - streamable-http
+canonical_mapping:
+  type_to_transport:
+    stdio: stdio
+    http: http
+    sse: streamable-http
+  httpUrl_to_url: true
+extra_fields:
+  - prompt
+  - sampling_enabled
+  - api_key_env
+  - api_key_header
+  - api_key_format
+  - startup_timeout_sec
+  - tool_timeout_sec
+```
+
+## Sources
+
+- Mistral Vibe README and docs pages for config path precedence
+- Upstream source definitions for MCP transport variants in `vibe/core/config`

__reports__/mistral_vibe/01-architecture_analysis_v0.md

@@ -0,0 +1,26 @@
+# Mistral Vibe Architecture Analysis
+
+## Model
+
+| Layer | Change |
+| --- | --- |
+| Unified model | Add Vibe-native fields and host enum |
+| Adapter | New `MistralVibeAdapter` to map canonical fields to Vibe TOML entries |
+| Strategy | New TOML strategy for `[[mcp_servers]]` read/write with key preservation |
+| Registries | Add adapter, strategy, backup/reporting, and fixture registration |
+| Tests | Extend generic adapter suites and add focused TOML strategy tests |
+
+## Integration Notes
+
+| Concern | Decision |
+| --- | --- |
+| Local vs global config | Prefer existing project-local file, otherwise global fallback |
+| Remote transport mapping | Canonical `type=sse` maps to Vibe `streamable-http` |
+| Cross-host sync | Accept canonical `type` and `httpUrl`, serialize to `transport` + `url` |
+| Non-MCP settings | Preserve other top-level TOML keys on write |
+
+## Assessment
+
+- **GO** — current adapter/strategy architecture already supports one more standalone TOML host.
+- No dependency installation is required.
+- Main regression surface is registry completeness and TOML round-tripping, covered by targeted tests.

__reports__/mistral_vibe/README.md

@@ -0,0 +1,12 @@
+# Mistral Vibe Reports
+
+## Status
+
+- Latest discovery: `00-parameter_analysis_v0.md`
+- Latest architecture analysis: `01-architecture_analysis_v0.md`
+- Current assessment: `GO`
+
+## Documents
+
+1. `00-parameter_analysis_v0.md` — upstream config path/schema discovery and host spec
+2. `01-architecture_analysis_v0.md` — integration plan, touched files, and go/no-go assessment

docs/articles/api/cli/mcp.md

@@ -24,6 +24,7 @@ This module provides handlers for:
 - codex: OpenAI Codex
 - lm-studio: LM Studio
 - gemini: Google Gemini
+- mistral-vibe: Mistral Vibe CLI coding agent
 
 ## Handler Functions
 

docs/articles/devs/architecture/mcp_backup_system.md

@@ -146,6 +146,7 @@ The system supports all MCP host platforms:
 | `cursor` | Cursor IDE MCP integration |
 | `lmstudio` | LM Studio MCP support |
 | `gemini` | Google Gemini MCP integration |
+| `mistral-vibe` | Mistral Vibe CLI coding agent |
 
 ## Performance Characteristics
 

docs/articles/devs/architecture/mcp_host_configuration.md

@@ -143,6 +143,7 @@ supported = registry.get_supported_hosts()  # List all hosts
 - `claude-desktop`, `claude-code`
 - `vscode`, `cursor`, `lmstudio`
 - `gemini`, `kiro`, `codex`
+- `mistral-vibe`
 
 ### BaseAdapter Protocol
 

docs/articles/users/CLIReference.md

@@ -1017,6 +1017,7 @@ Available MCP Host Platforms:
   claude-desktop      ✓ Available      /Users/user/.config/claude/...
   cursor              ✓ Available      /Users/user/.cursor/mcp.json
   vscode              ✗ Not Found      -
+  mistral-vibe        ✓ Available      /Users/user/.config/mistral/mcp.toml
 ```
 
 **Key Details**:
@@ -1039,6 +1040,8 @@ Available MCP host platforms:
     Config path: ~/.cursor/config.json
   vscode: ✗ Not detected
     Config path: ~/.vscode/config.json
+  mistral-vibe: ✓ Available
+    Config path: ~/.config/mistral/mcp.toml
 ```
 
 #### `hatch mcp discover servers`

docs/articles/users/MCPHostConfiguration.md

@@ -23,6 +23,7 @@ Hatch currently supports configuration for these MCP host platforms:
 - **Codex** - OpenAI Codex with MCP server configuration support
 - **LM Studio** - Local language model interface
 - **Gemini** - Google's AI development environment
+- **Mistral Vibe** - Mistral Vibe CLI coding agent
 
 ## Hands-on Learning
 

docs/articles/users/tutorials/04-mcp-host-configuration/01-host-platform-overview.md

@@ -57,6 +57,7 @@ Hatch currently supports configuration for these MCP host platforms:
 - [**Codex**](https://github.com/openai/codex) - OpenAI Codex with MCP server configuration support
 - [**LM Studio**](https://lmstudio.ai/) - Local language model interface
 - [**Gemini**](https://github.com/google-gemini/gemini-cli) - Google's AI Command Line Interface
+- [**Mistral Vibe**](https://mistral.ai/vibe) - Mistral Vibe CLI coding agent
 
 ## Configuration Management Workflow