[Security] Reapprove changed plugin artifacts and minimize MCP child environments

.env.example

@@ -1,92 +1,62 @@
-# CRITICAL
-# vLLM Studio Configuration
+# Local Studio Configuration
 # Copy this file to .env and modify as needed
+# Run `npm run doctor` to preflight your environment before first launch
 
 # =============================================================================
 # Controller Settings
 # =============================================================================
 
 # API server settings
-VLLM_STUDIO_HOST=127.0.0.1
-VLLM_STUDIO_PORT=8080
+LOCAL_STUDIO_HOST=127.0.0.1
+LOCAL_STUDIO_PORT=8080
 
-# Required when binding to a non-loopback host such as 0.0.0.0
-VLLM_STUDIO_API_KEY=
+# REQUIRED when LOCAL_STUDIO_HOST is non-loopback (e.g. 0.0.0.0) — the controller
+# refuses to start without it unless LOCAL_STUDIO_ALLOW_UNAUTHENTICATED=true.
+LOCAL_STUDIO_API_KEY=
 
-# Only set this to true for trusted local environments.
-# VLLM_STUDIO_ALLOW_UNAUTHENTICATED=false
+# Explicit opt-out of the API-key requirement on non-loopback hosts.
+# Only set this to true on trusted LANs.
+# LOCAL_STUDIO_ALLOW_UNAUTHENTICATED=false
 
 # Optional browser allowlist for direct controller access (comma-separated origins)
-# VLLM_STUDIO_CORS_ORIGINS=http://localhost:3000,http://127.0.0.1:3000
+# LOCAL_STUDIO_CORS_ORIGINS=http://localhost:3000,http://127.0.0.1:3000
+
+# --- Privileged-capability opt-ins (safe defaults are OFF) ---
+# Allow recipes to specify a raw launch_command/custom_command (arbitrary binary
+# execution). Leave off unless you rely on custom launch commands.
+# LOCAL_STUDIO_ALLOW_CUSTOM_LAUNCH_COMMAND=false
+# Default value of recipe trust_remote_code when a recipe omits it. Defaults to
+# true (models needing custom modeling code work out of the box). Set to false to
+# harden the default; recipes can still opt in per-recipe.
+# LOCAL_STUDIO_DEFAULT_TRUST_REMOTE_CODE=true
 
 # Inference backend port (where vLLM/SGLang runs)
-VLLM_STUDIO_INFERENCE_PORT=8000
+LOCAL_STUDIO_INFERENCE_PORT=8000
 
 # Enable mock inference mode (no external LLM required). Useful for local UI/E2E testing.
-# VLLM_STUDIO_MOCK_INFERENCE=true
+# LOCAL_STUDIO_MOCK_INFERENCE=true
 
 # =============================================================================
 # Paths
 # =============================================================================
 
-# Directory containing model weights
-VLLM_STUDIO_MODELS_DIR=/models
+# Directory containing model weights. Must be writable: the controller tries to
+# create it on startup and warns if it cannot.
+# LOCAL_STUDIO_MODELS_DIR=/models
 
 # Data directory for recipes and chat history
-VLLM_STUDIO_DATA_DIR=./data
+LOCAL_STUDIO_DATA_DIR=./data
 
 # =============================================================================
 # Backend-specific Settings (Optional)
 # =============================================================================
 
 # SGLang Python path (only needed if using SGLang backend)
-# VLLM_STUDIO_SGLANG_PYTHON=/path/to/sglang/venv/bin/python
-
-# TabbyAPI directory (only needed if using TabbyAPI/ExLlamaV3 backend)
-# VLLM_STUDIO_TABBY_API_DIR=/path/to/tabbyAPI
+# LOCAL_STUDIO_SGLANG_PYTHON=/path/to/sglang/venv/bin/python
 
 # llama.cpp server binary (only needed if llama-server is not on PATH)
-# VLLM_STUDIO_LLAMA_BIN=/usr/local/bin/llama-server
-
-# ExLLaMA v3 command template. Must resolve a runnable launch command.
-# Supports any binary with flags; include placeholders exactly as you want them to run.
-# Example: /opt/exllamav3/bin/exllama-server --model /models/my-model --port 8000
-# VLLM_STUDIO_EXLLAMAV3_COMMAND=exllama-server --host 127.0.0.1 --port 8000
+# LOCAL_STUDIO_LLAMA_BIN=/usr/local/bin/llama-server
 
 # Strict OpenAI model matching for /v1/chat/completions.
 # If true, only configured recipes are routable through the controller.
-# VLLM_STUDIO_STRICT_OPENAI_MODELS=false
-
-# Local filesystem fallback for agent file tools when Daytona fails or is unavailable.
-# Disabled by default. Set true only if you want model file ops to continue locally.
-# VLLM_STUDIO_AGENT_FS_LOCAL_FALLBACK=false
-
-# =============================================================================
-# LiteLLM Gateway (Optional - for API routing/format translation)
-# =============================================================================
-
-LITELLM_MASTER_KEY=dev-master-key
-VLLM_STUDIO_LITELLM_DATABASE_URL=postgresql://postgres:postgres@localhost:5432/litellm
-
-# Backend inference URL (used by LiteLLM to route requests)
-INFERENCE_API_BASE=http://localhost:8000/v1
-INFERENCE_API_KEY=dev-placeholder-key
-
-# =============================================================================
-# Frontend Settings
-# =============================================================================
-
-# LiteLLM URL for chat (defaults to localhost if not set)
-NEXT_PUBLIC_LITELLM_URL=http://localhost:4100
-
-# Frontend container user (helps write ./data on Linux hosts)
-VLLM_STUDIO_UID=1000
-VLLM_STUDIO_GID=1000
-
-# =============================================================================
-# Search Integration (Optional)
-# =============================================================================
-
-# Exa AI API key for web search (used in chat for research mode)
-# Get your API key from https://exa.ai
-EXA_API_KEY=your-exa-api-key-here
+# LOCAL_STUDIO_STRICT_OPENAI_MODELS=false

.githooks/commit-msg

@@ -0,0 +1,2 @@
+#!/bin/sh
+node scripts/check-conventional-commits.mjs --message-file "$1"

.githooks/pre-commit

@@ -0,0 +1,14 @@
+#!/bin/sh
+set -eu
+
+staged="$(git diff --cached --name-only)"
+
+if printf '%s\n' "$staged" | grep -qE '^(frontend|shared|tests/frontend)/'; then
+  echo "pre-commit: frontend staged lint/format + typecheck"
+  npm --prefix frontend run precommit
+fi
+
+if printf '%s\n' "$staged" | grep -q '^controller/'; then
+  echo "pre-commit: controller typecheck"
+  (cd controller && bun run typecheck)
+fi

.githooks/pre-push

@@ -0,0 +1,28 @@
+#!/bin/sh
+remote="$1"
+url="$2"
+
+while read local_ref local_sha remote_ref remote_sha
+do
+  if [ "$local_sha" = "0000000000000000000000000000000000000000" ]; then
+    continue
+  fi
+
+  if [ "$remote_sha" = "0000000000000000000000000000000000000000" ]; then
+    default_ref="$(git symbolic-ref --quiet --short "refs/remotes/$remote/HEAD" 2>/dev/null || printf '%s/main' "$remote")"
+    base_sha="$(git merge-base "$default_ref" "$local_sha" 2>/dev/null || true)"
+    if [ -n "$base_sha" ]; then
+      range="$base_sha..$local_sha"
+    else
+      range="$local_sha"
+    fi
+  else
+    range="$remote_sha..$local_sha"
+  fi
+
+  echo "Checking conventional commits for $local_ref -> $remote/$remote_ref ($url)"
+  node scripts/check-conventional-commits.mjs --range "$range" || exit 1
+done
+
+echo "pre-push: frontend quality gate"
+npm --prefix frontend run check:quality

.github/CODEOWNERS

@@ -6,12 +6,9 @@
 # Application-specific owners
 /controller/ @0xSero
 /frontend/ @0xSero
-/cli/ @0xSero
 
 # Configuration
 docker-compose.yml @0xSero
-config/ @0xSero
 
 # Documentation
 *.md @0xSero
-docs/ @0xSero

.github/ISSUE_TEMPLATE/bug_report.md

@@ -2,7 +2,7 @@
 name: Bug report
 about: Create a report to help us improve
 title: '[BUG] '
-labels: bug
+labels: "Type: Bug"
 assignees: ''
 ---
 
@@ -25,8 +25,8 @@ A clear and concise description of what you expected to happen.
 ## Environment
 
 - **OS**: [e.g. Ubuntu 22.04, macOS 14.0]
-- **vLLM Studio Version**: [e.g. 0.3.1]
-- **Backend**: [vLLM / SGLang / TabbyAPI]
+- **Local Studio Version**: [e.g. 0.3.1]
+- **Backend**: [vLLM / SGLang / llama.cpp / MLX]
 - **Model**: [e.g. Llama-3-8B-Instruct]
 
 ## Logs

.github/ISSUE_TEMPLATE/feature_request.md

@@ -2,7 +2,7 @@
 name: Feature request
 about: Suggest an idea for this project
 title: '[FEATURE] '
-labels: enhancement
+labels: "Type: Feature"
 assignees: ''
 ---
 

.github/labels.yml

@@ -1,4 +1,3 @@
-# CRITICAL
 # Repository Issue Labels
 
 # Priority Labels

.github/pull_request_template.md

@@ -1,57 +1,22 @@
-## Description
+## Summary
 
-Briefly describe the changes in this PR.
+Describe the change in 1-3 sentences.
 
-## Type of Change
+## Validation
 
-- [ ] Bug fix (non-breaking change which fixes an issue)
-- [ ] New feature (non-breaking change which adds functionality)
-- [ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)
-- [ ] Documentation update
-- [ ] Refactoring (no functional changes)
-- [ ] CI/CD changes
-- [ ] Performance improvements
+List the commands you ran:
 
-## Related Issues
+```bash
+# example
+npm --prefix frontend run lint
+npm --prefix frontend run typecheck
+npm --prefix frontend test
+```
 
-Fixes #
-Related to #
+## UI changes
 
-## Testing
+Attach screenshots or recordings if this changes visible behavior.
 
-Describe the testing performed for this PR:
-- [ ] Unit tests pass (`bun test` / `npm test` / `pytest`)
-- [ ] Integration tests pass (Playwright)
-- [ ] Manual testing performed
-- [ ] All linting passes (`bun run lint` / `npm run lint`)
-- [ ] Type checking passes (`bun run typecheck`)
-- [ ] Dead code check passes (`bun run check`)
+## Risks / rollout notes
 
-Include specific test scenarios or commands used.
-
-## Checklist
-
-- [ ] My code follows the style guidelines of this project
-- [ ] I have performed a self-review of my code
-- [ ] I have commented my code, particularly in hard-to-understand areas
-- [ ] I have made corresponding changes to the documentation
-- [ ] My changes generate no new warnings
-- [ ] I have tested my changes locally
-- [ ] I have updated AGENTS.md if needed (for architectural changes)
-- [ ] I have added appropriate labels (Priority, Type, Area)
-- [ ] No new security vulnerabilities introduced
-
-## Performance & Security
-
-- [ ] Performance impact considered (no significant regressions)
-- [ ] Dependencies audited (no known vulnerabilities)
-- [ ] Secrets not exposed
-- [ ] Error handling implemented appropriately
-
-## Screenshots (if applicable)
-
-Add screenshots to help explain your changes.
-
-## Additional Notes
-
-Any additional information or context for reviewers.
+Call out migrations, compatibility concerns, deployment steps, or follow-up work.