mirror of https://github.com/ruvnet/RuVector.git synced 2026-05-25 23:24:03 +00:00

rUv 930fca916f feat(sse): decouple SSE to mcp.pi.ruv.io proxy + Claude Code source research

SSE Proxy Decoupling (ADR-130):
- Fix ruvbrain-sse proxy: proper MCP handshake, session creation, drain polling
- Fix internal queue endpoints: session_create keeps receiver, drain returns buffered messages
- Add response_queues to AppState for SSE proxy communication
- Skip sparsifier for >5M edge graphs (was crashing on 16M edges)
- Add SSE_DISABLED/MAX_SSE env vars for configurable connection limits
- Route SSE to dedicated mcp.pi.ruv.io subdomain (Cloudflare CNAME)
- Serve SSE at root / path on proxy (no /sse needed)
- Update all references from pi.ruv.io/sse to mcp.pi.ruv.io
- Fix Dockerfile consciousness crate build (feature/version mismatches)

Claude Code CLI Source Research (ADR-133):
- 19 research documents analyzing Claude Code internals (3000+ lines)
- Decompiler script + RVF corpus builder for all major versions
- Binary RVF containers for v0.2, v1.0, v2.0, v2.1 (300-2068 vectors each)
- Call graphs, class hierarchies, state machines from minified source

Integration Strategy (ADR-134):
- 6-tier integration plan: WASM MCP, agents, hooks, cache, SDK, plugin
- Integration guide with architecture diagrams and performance targets

Co-Authored-By: claude-flow <ruv@ruv.net>

2026-04-02 23:39:56 +00:00

8.4 KiB

Raw Permalink Blame History

Claude Code CLI: Configuration and Environment

Configuration Hierarchy

Settings are loaded from multiple sources with increasing priority:

1. Built-in defaults (lowest)
2. User settings:        ~/.claude/settings.json
3. Project settings:     .claude/settings.json
4. Local settings:       .claude/settings.local.json
5. Managed settings:     Enterprise/team (highest for locked fields)
6. CLI flags:            --model, --permission-mode, etc. (runtime override)
7. Environment variables: ANTHROPIC_MODEL, etc. (runtime override)

--setting-sources controls which sources load: user, project, local.

Settings Schema (76 Properties)

Model Configuration

Setting	Type	Description
`model`	string	Override default model
`advisorModel`	string	Model for advisor tool
`availableModels`	array	Allowlisted models (e.g., "opus", "sonnet")
`modelOverrides`	object	Map Anthropic model IDs to provider-specific IDs
`effortLevel`	string	Persisted effort level
`fastMode`	boolean	Enable fast mode
`fastModePerSessionOptIn`	boolean	Fast mode does not persist across sessions
`alwaysThinkingEnabled`	boolean	Control thinking/reasoning

Agent Configuration

Setting	Type	Description
`agent`	string	Default agent for main thread
`permissions`	object	Tool permission rules (allow/deny/ask)
`hooks`	object	Hook event handlers
`env`	object	Environment variables for sessions

MCP Configuration

Setting	Type	Description
`allowedMcpServers`	array	Enterprise MCP server allowlist
`deniedMcpServers`	array	Enterprise MCP server denylist
`enableAllProjectMcpServers`	boolean	Auto-approve project MCP servers
`enabledMcpjsonServers`	array	Approved .mcp.json servers
`disabledMcpjsonServers`	array	Rejected .mcp.json servers
`allowManagedMcpServersOnly`	boolean	Lock to managed servers only

Plugin/Marketplace

Setting	Type	Description
`enabledPlugins`	object	Active plugins (plugin-id@marketplace-id)
`pluginConfigs`	object	Per-plugin configuration
`blockedMarketplaces`	array	Blocked marketplace sources
`extraKnownMarketplaces`	object	Additional marketplaces
`allowedChannelPlugins`	array	Channel plugin allowlist
`channelsEnabled`	boolean	Enable channel notifications
`pluginTrustMessage`	string	Custom trust warning message

Session/Memory

Setting	Type	Description
`autoMemoryEnabled`	boolean	Enable auto-memory
`autoMemoryDirectory`	string	Custom memory storage path
`autoDreamEnabled`	boolean	Background memory consolidation
`cleanupPeriodDays`	integer	Transcript retention (default: 30)
`autoCompactWindow`	integer	Auto-compact window size
`plansDirectory`	string	Custom plan file directory

Security/Enterprise

Setting	Type	Description
`allowManagedHooksOnly`	boolean	Lock to managed hooks
`allowManagedPermissionRulesOnly`	boolean	Lock to managed permissions
`disableAllHooks`	boolean	Disable all hooks
`disableAutoMode`	string	Disable auto permission mode
`forceLoginMethod`	string	Force "claudeai" or "console" auth
`forceLoginOrgUUID`	string/array	Required org UUID for OAuth
`minimumVersion`	string	Prevent downgrades

Display/UX

Setting	Type	Description
`language`	string	Response language preference
`outputStyle`	string	Output style for responses
`prefersReducedMotion`	boolean	Reduce animations
`feedbackSurveyRate`	number	Survey probability (0-1)
`attribution`	object	Commit/PR attribution text
`includeGitInstructions`	boolean	Include git workflow prompts
`promptSuggestionEnabled`	boolean	Enable prompt suggestions
`fileSuggestion`	object	@ mention file suggestions

Remote

Setting	Type	Description
`remote`	object	Remote session configuration
`defaultShell`	string	Default shell (defaults to bash)
`apiKeyHelper`	string	Script path for auth values
`awsAuthRefresh`	string	AWS auth refresh script
`awsCredentialExport`	string	AWS credential export script
`gcpAuthRefresh`	string	GCP auth refresh command
`otelHeadersHelper`	string	OTEL headers script

Environment Variables (498 Recognized)

Critical Anthropic Variables

Variable	Purpose
`ANTHROPIC_API_KEY`	Primary API key
`ANTHROPIC_AUTH_TOKEN`	Auth token
`ANTHROPIC_BASE_URL`	Custom API base URL
`ANTHROPIC_MODEL`	Model override
`ANTHROPIC_BETAS`	Beta feature flags
`ANTHROPIC_CUSTOM_HEADERS`	Custom API headers

Provider Variables

Variable	Purpose
`ANTHROPIC_BEDROCK_BASE_URL`	Bedrock endpoint
`CLAUDE_CODE_USE_BEDROCK`	Enable Bedrock
`CLAUDE_CODE_USE_VERTEX`	Enable Vertex AI
`CLAUDE_CODE_USE_FOUNDRY`	Enable Azure Foundry
`CLAUDE_CODE_USE_ANTHROPIC_AWS`	Enable Anthropic AWS
`AWS_ACCESS_KEY_ID`	AWS credentials
`AWS_SECRET_ACCESS_KEY`	AWS credentials
`AWS_SESSION_TOKEN`	AWS session
`AWS_REGION`	AWS region
`ANTHROPIC_VERTEX_PROJECT_ID`	GCP project
`ANTHROPIC_FOUNDRY_BASE_URL`	Azure endpoint
`ANTHROPIC_FOUNDRY_API_KEY`	Azure key

Model Configuration Variables

Variable	Purpose
`ANTHROPIC_DEFAULT_SONNET_MODEL`	Default Sonnet model ID
`ANTHROPIC_DEFAULT_OPUS_MODEL`	Default Opus model ID
`ANTHROPIC_DEFAULT_HAIKU_MODEL`	Default Haiku model ID
`ANTHROPIC_SMALL_FAST_MODEL`	Small/fast model for subtasks
`CLAUDE_CODE_SUBAGENT_MODEL`	Model for subagents

Feature Flags (CLAUDE_CODE_*)

Over 120 CLAUDE_CODE_* variables control features:

Disable flags: CLAUDE_CODE_DISABLE_*

_AUTO_MEMORY, _BACKGROUND_TASKS, _CLAUDE_MDS, _CRON
_FAST_MODE, _FILE_CHECKPOINTING, _GIT_INSTRUCTIONS
_THINKING, _VIRTUAL_SCROLL, _MOUSE, _TERMINAL_TITLE
_ATTACHMENTS, _EXPERIMENTAL_BETAS, _FEEDBACK_SURVEY
_1M_CONTEXT, _ADAPTIVE_THINKING, _POLICY_SKILLS

Enable flags: CLAUDE_CODE_ENABLE_*

_CFC, _FINE_GRAINED_TOOL_STREAMING, _PROMPT_SUGGESTION
_SDK_FILE_CHECKPOINTING, _TASKS, _TELEMETRY
_TOKEN_USAGE_ATTACHMENT, _XAA

Configuration: CLAUDE_CODE_*

_EFFORT_LEVEL, _MAX_OUTPUT_TOKENS, _MAX_RETRIES
_SHELL, _SHELL_PREFIX, _TMPDIR
_GLOB_TIMEOUT_SECONDS, _GLOB_HIDDEN, _GLOB_NO_IGNORE
_SCROLL_SPEED, _IDLE_THRESHOLD_MINUTES
_OAUTH_TOKEN, _OAUTH_CLIENT_ID, _OAUTH_SCOPES

Bash Sandbox Variables

Variable	Purpose
`BASH_MAX_OUTPUT_LENGTH`	Max Bash tool output
`CLAUDE_CODE_BUBBLEWRAP`	Bubblewrap sandbox config
`CLAUDE_CODE_BASH_SANDBOX_SHOW_INDICATOR`	Show sandbox indicator
`CLAUDE_CODE_BASH_MAINTAIN_PROJECT_WORKING_DIR`	Maintain cwd

Claude Home Directory Structure

~/.claude/
  settings.json           # User settings
  keybindings.json        # Custom keybindings
  CLAUDE.md               # Global user instructions
  scheduled_tasks.json    # Cron-like tasks
  agents/                 # Custom agent definitions
  commands/               # Custom slash commands
  debug/                  # Debug logs
  local/                  # Local data
    claude                # Local state
    node_modules/         # Cached modules
  plans/                  # Saved plans
  plugins/                # Installed plugins
    data/                 # Plugin data
  projects/               # Per-project data
    <hash>/               # Project-specific
      memory/MEMORY.md    # Auto-memory
  remote/                 # Remote session data
  rules/                  # Custom rules

Project Configuration Files

.claude/
  settings.json           # Project settings
  settings.local.json     # Local overrides (gitignored)
  agents/                 # Project-specific agents
  skills/
    deploy/SKILL.md       # Skill definitions

.mcp.json                 # MCP server definitions
.claudeignore             # Files to ignore
CLAUDE.md                 # Project instructions

8.4 KiB Raw Permalink Blame History