fix: correct config-reference schema and v5 frontmatter artifacts

- Rewrite config-reference.md to match actual MaxsimConfig interface
- Update commands-todos.md frontmatter id to commands-quick
- Remove non-existent --existing flag from commands-milestone.md
- Fix --existing references in codebase-mapping.md and commands-core.md

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: maystudios

packages/website/src/content/docs/codebase-mapping.md

@@ -6,17 +6,17 @@ group: Advanced
 
 Codebase mapping analyzes an existing codebase using parallel mapper agents. It produces structured analysis that subsequent planning agents use as context when working in the project.
 
-Codebase mapping runs automatically as part of the `/maxsim:init --existing` workflow. When you onboard MaxsimCLI to a project that already has code, the init process dispatches mapper agents to understand the codebase before any planning begins.
+Codebase mapping runs automatically as part of the `/maxsim:init` workflow when MaxsimCLI detects an existing codebase. When you onboard MaxsimCLI to a project that already has code, the init process dispatches mapper agents to understand the codebase before any planning begins.
 
 {% codeblock language="bash" %}
 # Initialize MaxsimCLI in an existing project — triggers codebase mapping automatically
-/maxsim:init --existing
+/maxsim:init
 {% /codeblock %}
 
 Multiple codebase-mapper agents run in parallel, each covering a different area of the codebase. One covers data models, another covers API routes, another covers frontend components, another covers infrastructure. Their outputs are synthesized into a unified analysis stored in the GitHub Wiki and cached in `.claude/agent-memory/` for fast agent access.
 
 The codebase analysis is loaded automatically by phase-researcher agents. When planning a phase in an existing codebase, the researcher reads the analysis to understand existing patterns, conventions, and potential integration points instead of re-discovering them from scratch.
 
 {% callout type="note" %}
-Codebase mapping is not a standalone command. It is triggered as part of project initialization with `/maxsim:init --existing`. If you need to re-map the codebase after significant changes, run the init workflow again.
+Codebase mapping is not a standalone command. It is triggered as part of project initialization with `/maxsim:init` when an existing codebase is detected. If you need to re-map the codebase after significant changes, run the init workflow again.
 {% /callout %}

packages/website/src/content/docs/commands-core.md

@@ -4,5 +4,5 @@ title: Core Commands
 group: Commands Reference
 ---
 
-{% doctable headers=["Command", "Description", "Key Flags"] rows=[["/maxsim:debug", "Start a structured debugging session", "—"], ["/maxsim:debug-loop [symptom]", "Autonomous bug hunting loop for a given symptom", "—"], ["/maxsim:execute", "Execute a phase plan with parallel agents", "--worktrees, --no-worktrees"], ["/maxsim:fix-loop [cmd]", "Autonomous error repair loop for a failing command", "—"], ["/maxsim:go", "Auto-dispatch: detects state, runs right workflow", "—"], ["/maxsim:help", "Show available commands and usage", "—"], ["/maxsim:improve [metric]", "Autonomous optimization loop targeting a metric", "—"], ["/maxsim:init", "Initialize project, onboard existing, manage milestones", "--existing"], ["/maxsim:plan", "Plan a phase (discussion, research, planning)", "--force-research, --skip-verify"], ["/maxsim:progress", "Show current phase/milestone progress", "—"], ["/maxsim:quick", "Run a quick standalone task", "—"], ["/maxsim:security [scope]", "Security audit of the given scope (read-only)", "—"], ["/maxsim:settings", "View or modify MaxsimCLI configuration", "—"]] %}
+{% doctable headers=["Command", "Description", "Key Flags"] rows=[["/maxsim:debug", "Start a structured debugging session", "—"], ["/maxsim:debug-loop [symptom]", "Autonomous bug hunting loop for a given symptom", "—"], ["/maxsim:execute", "Execute a phase plan with parallel agents", "--worktrees, --no-worktrees"], ["/maxsim:fix-loop [cmd]", "Autonomous error repair loop for a failing command", "—"], ["/maxsim:go", "Auto-dispatch: detects state, runs right workflow", "—"], ["/maxsim:help", "Show available commands and usage", "—"], ["/maxsim:improve [metric]", "Autonomous optimization loop targeting a metric", "—"], ["/maxsim:init", "Initialize project, onboard existing, manage milestones", "—"], ["/maxsim:plan", "Plan a phase (discussion, research, planning)", "--force-research, --skip-verify"], ["/maxsim:progress", "Show current phase/milestone progress", "—"], ["/maxsim:quick", "Run a quick standalone task", "—"], ["/maxsim:security [scope]", "Security audit of the given scope (read-only)", "—"], ["/maxsim:settings", "View or modify MaxsimCLI configuration", "—"]] %}
 {% /doctable %}

packages/website/src/content/docs/commands-milestone.md

@@ -12,15 +12,12 @@ Milestone lifecycle is managed through `/maxsim:init`. This single entry point h
 
 ### Operations available through `/maxsim:init`
 
-{% doctable headers=["Operation", "Description"] rows=[["new-project", "Initialize a new project — Creates GitHub repo, Project Board, Milestones with phase issues."], ["init-existing", "Onboard an existing codebase into MaxsimCLI (use --existing flag)"], ["new-milestone", "Add a new milestone to the Project Board with placeholder phase issues"], ["complete-milestone", "Archive milestone phases and advance to the next milestone"]] %}
+{% doctable headers=["Operation", "Description"] rows=[["new-project", "Initialize a new project — Creates GitHub repo, Project Board, Milestones with phase issues."], ["init-existing", "Onboard an existing codebase into MaxsimCLI (auto-detected when the repo already has code)"], ["new-milestone", "Add a new milestone to the Project Board with placeholder phase issues"], ["complete-milestone", "Archive milestone phases and advance to the next milestone"]] %}
 {% /doctable %}
 
 {% codeblock language="bash" %}
-# Initialize a new project
+# Initialize a new project or onboard an existing one (auto-detected)
 /maxsim:init
-
-# Onboard an existing codebase
-/maxsim:init --existing
 {% /codeblock %}
 
 Milestone auditing and gap planning are handled as part of the init workflow. When completing a milestone, MaxsimCLI audits deliverables against original requirements, identifies gaps, and creates fix phases before archiving.

packages/website/src/content/docs/commands-todos.md

@@ -1,5 +1,5 @@
 ---
-id: commands-todos
+id: commands-quick
 title: Quick Tasks
 group: Commands Reference
 ---

packages/website/src/content/docs/config-reference.md

@@ -11,6 +11,8 @@ Place `.claude/maxsim/config.json` to customize MaxsimCLI behavior per-project.
   "version": "6.0.0",
   "execution": {
     "model_profile": "balanced",
+    "competitive_enabled": false,
+    "model_overrides": {},
     "parallelism": {
       "max_agents_per_wave": 3,
       "max_retries": 3,
@@ -24,16 +26,34 @@ Place `.claude/maxsim/config.json` to customize MaxsimCLI behavior per-project.
     }
   },
   "worktrees": {
-    "basePath": ".maxsim-worktrees/",
     "auto_cleanup": true,
-    "branch_prefix": "maxsim/"
+    "branch_prefix": "maxsim/",
+    "path_template": ".claude/worktrees/agent-{id}/",
+    "branch_template": "maxsim/phase-{N}-task-{id}"
   },
   "automation": {
     "auto_commit_on_success": true,
-    "conventional_commits": true
+    "conventional_commits": true,
+    "co_author": "Co-Authored-By: Claude <noreply@anthropic.com>"
+  },
+  "github": {
+    "projectName": "",
+    "auto_push": true
+  },
+  "hooks": {
+    "enabled": true
+  },
+  "workflow": {
+    "research": true,
+    "plan_checker": true,
+    "verifier": true,
+    "auto_advance": false
+  },
+  "git": {
+    "branching_strategy": "phase"
   }
 }
 {% /codeblock %}
 
-{% doctable headers=["Key", "Type", "Default", "Description"] rows=[["execution.model_profile", "string", "balanced", "Active model profile for all agents (quality, balanced, budget)"], ["execution.parallelism.max_agents_per_wave", "number", "3", "Maximum number of agents that can run concurrently in a wave"], ["execution.parallelism.max_retries", "number", "3", "Maximum retry attempts when an agent fails"], ["execution.parallelism.competition_strategy", "string", "standard", "Strategy for competitive implementation (none, quick, standard, deep)"], ["execution.verification.strict_mode", "boolean", "true", "Require all verification gates to pass before marking execution complete"], ["execution.verification.gates", "array", "(all)", "List of verification gates to enforce (tests_pass, build_succeeds, lint_clean, spec_compliance, code_review)"], ["execution.verification.require_code_review", "boolean", "true", "Require code review gate during verification"], ["execution.verification.auto_resolve_conflicts", "boolean", "true", "Automatically resolve minor conflicts during parallel execution"], ["worktrees.basePath", "string", ".maxsim-worktrees/", "Directory used for git worktree isolation during parallel execution"], ["worktrees.auto_cleanup", "boolean", "true", "Remove worktrees automatically after agent completion"], ["worktrees.branch_prefix", "string", "maxsim/", "Prefix for worktree branch names"], ["automation.auto_commit_on_success", "boolean", "true", "Automatically commit changes when a task completes successfully"], ["automation.conventional_commits", "boolean", "true", "Enforce conventional commit message format"]] %}
+{% doctable headers=["Key", "Type", "Default", "Description"] rows=[["execution.model_profile", "string", "balanced", "Active model profile for all agents (quality, balanced, budget)"], ["execution.competitive_enabled", "boolean", "false", "Enable competitive implementation where multiple agents solve the same task"], ["execution.model_overrides", "object", "{}", "Per-agent model overrides (keys: planner, executor, researcher, verifier; values: haiku, sonnet, opus)"], ["execution.parallelism.max_agents_per_wave", "number", "3", "Maximum number of agents that can run concurrently in a wave"], ["execution.parallelism.max_retries", "number", "3", "Maximum retry attempts when an agent fails"], ["execution.parallelism.competition_strategy", "string", "standard", "Strategy for competitive implementation (none, quick, standard, deep)"], ["execution.verification.strict_mode", "boolean", "true", "Require all verification gates to pass before marking execution complete"], ["execution.verification.gates", "array", "(all)", "List of verification gates to enforce (tests_pass, build_succeeds, lint_clean, spec_compliance, code_review)"], ["execution.verification.require_code_review", "boolean", "true", "Require code review gate during verification"], ["execution.verification.auto_resolve_conflicts", "boolean", "true", "Automatically resolve minor conflicts during parallel execution"], ["worktrees.auto_cleanup", "boolean", "true", "Remove worktrees automatically after agent completion"], ["worktrees.branch_prefix", "string", "maxsim/", "Prefix for worktree branch names"], ["worktrees.path_template", "string", ".claude/worktrees/agent-{id}/", "Path template for worktree directories ({id} is replaced with the agent id)"], ["worktrees.branch_template", "string", "maxsim/phase-{N}-task-{id}", "Branch name template for worktrees ({N} = phase number, {id} = task id)"], ["automation.auto_commit_on_success", "boolean", "true", "Automatically commit changes when a task completes successfully"], ["automation.conventional_commits", "boolean", "true", "Enforce conventional commit message format"], ["automation.co_author", "string", "Co-Authored-By: Claude <noreply@anthropic.com>", "Co-author trailer appended to commit messages"], ["github.projectName", "string", "\"\"", "GitHub Project Board name for issue tracking"], ["github.auto_push", "boolean", "true", "Automatically push commits to the remote after successful execution"], ["hooks.enabled", "boolean", "true", "Enable the MaxsimCLI hook system for lifecycle events"], ["workflow.research", "boolean", "true", "Run research agent during planning phases"], ["workflow.plan_checker", "boolean", "true", "Run plan-checker agent to validate phase plans before execution"], ["workflow.verifier", "boolean", "true", "Run verifier agent after execution to validate results"], ["workflow.auto_advance", "boolean", "false", "Automatically advance to the next phase after successful execution"], ["git.branching_strategy", "string", "phase", "Git branching strategy (none, phase, milestone)"]] %}
 {% /doctable %}