@hailer/mcp 1.2.0 → 2.0.0-beta.1

package/.claude/commands/help:faq.md

@@ -1,79 +0,0 @@
----
-description: Frequently asked questions for Hailer MCP
----
-
-# FAQ
-
-Display frequently asked questions to the user.
-
-## Output
-
-```
-╭─────────────────────────────────────────╮
-│     HAILER MCP - FAQ                    │
-╰─────────────────────────────────────────╯
-
-Q: Why is Claude failing at Hailer tasks?
-A: Check two things:
-   1. Run /mcp-list to verify Claude is connected to the Hailer MCP
-   2. Make sure you're running Claude from the correct project folder
-      (the one with CLAUDE.md and .claude/ folder)
-
-Q: Why do I need to restart Claude Code after installing a plugin?
-A: Claude Code reloads configuration at startup. New skills and hooks
-   won't be active until restart. Use 'claude -c' to keep context.
-
-Q: What's the difference between /plugin install and /install-plugin?
-A: /plugin install - Claude's built-in plugin system (global)
-   /install-plugin - Our local marketplace system (per-project)
-
-Q: Can I create my own skills?
-A: Yes! Add a SKILL.md file to .claude/skills/<name>/ following the
-   skill structure. See existing skills for examples.
-
-Q: How do I publish my skill or hook to the marketplace?
-A: Use /publish-plugin. It creates a PR that gets reviewed.
-   See /help plugins for details.
-
-Q: Why does Claude spawn sub-agents instead of doing work directly?
-A: Sub-agents with focused skills produce better results for domain-
-   specific tasks by loading only the relevant context.
-
-Q: What happens if a sub-agent fails?
-A: The orchestrator receives an error and reports the failure.
-   It should identify root cause and fix it, never apply bandaids.
-
-Q: How does Claude decide which skills to use?
-A: Check the skill catalog in CLAUDE.md. Skills are loaded based on
-   the task domain (workspace config, app building, testing, etc.).
-
-Q: Can I use MCP tools directly without agents?
-A: Yes! Tools like list_workflows, create_activity, etc. are
-   available directly. Agents just provide structured workflows.
-
-Q: Where is configuration stored?
-A: - Skills: .claude/skills/
-   - Hooks: .claude/hooks/
-   - Commands: .claude/commands/
-   - Settings: .claude/settings.json
-   - Instructions: CLAUDE.md
-
-Q: How do I update an installed plugin?
-A: Uninstall and reinstall:
-   /uninstall-plugin <name>
-   /install-plugin <name>
-
-Q: What's the difference between /autoplan and /plan-workspace?
-A: /autoplan - Quick planning, creates DEVELOPMENT.md + PRDs
-   /plan-workspace - Thorough 10-phase discovery for complex workspaces
-   Use /autoplan for features, /plan-workspace for new projects.
-
-Q: How do PRDs and DEVELOPMENT.md work together?
-A: DEVELOPMENT.md is the big picture (roadmap, status, decisions).
-   PRDs (docs/prd-*.md) are detailed requirements for each feature.
-   DEVELOPMENT.md roadmap links to individual PRDs.
-
-MORE HELP:
-  /help plugins  - Plugin system
-  /help agents   - Agent usage
-```

package/.claude/commands/help:plugins.md

@@ -1,50 +0,0 @@
----
-description: Plugin system help for Hailer MCP
----
-
-# Plugin System Help
-
-Display plugin system instructions to the user.
-
-## Output
-
-```
-HAILER MCP - PLUGIN SYSTEM
-
-The plugin system lets you install community agents from the marketplace.
-
-COMMANDS:
-
-  /marketplace-setup
-    Clone or pull the marketplace repo to ./hailer-marketplace
-    Only needed once. Commands will prompt you if missing.
-
-  /list-plugins
-    List all available plugins (auto-pulls latest).
-
-  /install-plugin <name>
-    Install a plugin to .claude/ folder.
-    Example: /install-plugin my-custom-plugin
-
-  /uninstall-plugin <name>
-    Remove a plugin from .claude/ folder.
-
-  /publish-plugin
-    Publish your plugin to marketplace (creates PR).
-
-PLUGIN TYPES:
-
-  Agents   AI specialists (.claude/agents/)
-  Skills   On-demand docs (.claude/skills/)
-  Hooks    Event scripts (.claude/hooks/)
-
-WORKFLOW:
-
-  1. /list-plugins            # See what's available
-  2. /install-plugin <name>   # Install what you need
-  3. Restart: claude -c       # Load new agents
-
-SEE ALSO:
-  /help agents   - How to use installed agents
-  /help faq      - Common questions
-```

package/.claude/commands/help:skills.md

@@ -1,87 +0,0 @@
----
-description: Skills reference for Hailer MCP
----
-
-# Skills Reference
-
-Display skills reference to the user.
-
-## Output
-
-```
-╭─────────────────────────────────────────╮
-│     HAILER MCP - SKILLS REFERENCE       │
-╰─────────────────────────────────────────╯
-
-Skills are reusable knowledge files (.claude/skills/<name>/SKILL.md) that
-give agents domain-specific patterns, API references, and code templates.
-
-HOW SKILLS WORK:
-
-  Auto-injected: Agents declare skills in their frontmatter. The
-  SubagentStart hook injects them automatically when the agent spawns.
-
-  On-demand: The orchestrator tells an agent to load a skill via the
-  Skill tool. Load skills before spawning sub-agents or doing domain-specific work.
-
-─────────────────────────────────────────────────────────────────────────
-
-SDK SKILLS (core patterns):
-  sdk-activity-patterns    - Field value formats, dates, users, links
-  sdk-document-templates   - PDF/CSV template structure + generation
-  sdk-function-fields      - Calculated fields in Hailer workflows
-  sdk-insight-queries      - SQL query patterns for insights, cross-workflow JOINs
-  sdk-ws-config-skill      - Workflow config: fields, phases, schemas, type generation, project setup
-
-HAILER PLATFORM SKILLS:
-  hailer-api-client        - @hailer/cli auth, requests, signals, direct HTTP API endpoints
-  hailer-app-builder       - Build apps with @hailer/app-sdk
-  hailer-apps-pictures     - Fetch and display images from Hailer
-  hailer-design-system     - Chakra UI theme, colors, icons, components, UI/UX patterns
-  hailer-monolith-automations - Webhooks, scheduled jobs in monolith, phase cascade bots
-  hailer-permissions-system - Users, teams, access control
-  hailer-project-protocol  - Full SDK workflow reference
-
-INTEGRATION SKILLS:
-  integration-patterns     - Activity movers, webhooks, SCIM, Kafka
-  zapier-hailer-patterns   - Zapier triggers, actions, Zap configs
-TOOL & AGENT SKILLS:
-  agent-structure          - Template for creating new agents
-  delegation-routing       - Full routing table + chaining patterns
-  tool-builder             - Building MCP tools for Hailer server
-  mcp-direct-tools         - MCP tools not assigned to agents
-
-QUALITY & FIX SKILLS (auto-loaded on errors):
-  tool-parameter-usage     - Extract IDs, correct parameter formats
-  tool-response-verification - Verify MCP results before reporting
-  optional-parameters      - Omit optional params vs empty values
-  json-only-output         - Fix agents adding prose after JSON
-  testing-patterns         - vitest, playwright, build verification
-
-REFERENCE SKILLS:
-  api-documentation-patterns - API endpoint documentation standards
-  chrome-mcp-reference     - Browser automation reference
-  lsp-setup                - Language Server Protocol setup
-  publish-hailer-app       - App publishing to production
-
-─────────────────────────────────────────────────────────────────────────
-
-ON-DEMAND SKILL LOADING:
-
-  Tell the orchestrator which skill a sub-agent needs:
-  "Load hailer-apps-pictures for the app builder sub-agent"
-  → Sub-agent gets image handling patterns
-
-  "Load sdk-insight-queries for this SQL query"
-  → Sub-agent gets insight patterns including JOIN syntax for cross-workflow data
-
-  Skill content is loaded by the orchestrator and pasted directly
-  into the sub-agent prompt — agents never load skills themselves.
-
-─────────────────────────────────────────────────────────────────────────
-
-SEE ALSO:
-  /help agents   - Which agents use which skills
-  /help commands - All slash commands
-  /help tools    - MCP tools reference
-```

package/.claude/commands/help:tools.md

@@ -1,75 +0,0 @@
----
-description: MCP tools reference for Hailer MCP
----
-
-# MCP Tools Reference
-
-Display MCP tools reference to the user.
-
-## Output
-
-```
-╭─────────────────────────────────────────╮
-│      HAILER MCP - TOOLS REFERENCE       │
-╰─────────────────────────────────────────╯
-
-MCP tools provide direct access to Hailer APIs. These are used by
-agents but can also be called directly.
-
-WORKFLOW TOOLS:
-  list_workflows          - List all workflows in workspace
-  list_workflows_minimal  - Compact workflow list with counts
-  list_workflow_phases    - Get phases for a workflow
-  get_workflow_schema     - Get field definitions for a workflow
-  install_workflow        - Create workflow from template
-
-ACTIVITY TOOLS:
-  list_activities         - List activities in a workflow phase
-  show_activity_by_id     - Get single activity details
-  create_activity         - Create new activity (single or bulk)
-  update_activity         - Update activity (single or bulk)
-  count_activities        - Count activities in workflow
-
-DISCUSSION TOOLS:
-  list_my_discussions        - List discussions you're in
-  fetch_discussion_messages  - Read messages from discussion
-  add_discussion_message     - Post to a discussion
-  join_discussion            - Join a discussion
-  leave_discussion           - Leave a discussion
-  invite_discussion_members  - Invite users to discussion
-
-INSIGHT TOOLS:
-  list_insights      - List all insights
-  create_insight     - Create SQL-like insight
-  preview_insight    - Test insight query
-  get_insight_data   - Execute insight and get results
-  update_insight     - Modify existing insight
-
-APP TOOLS:
-  list_apps              - List apps in workspace
-  create_app             - Create app entry
-  update_app             - Update app properties
-  scaffold_hailer_app    - Generate app from template
-  publish_hailer_app     - Deploy app to production
-
-FILE TOOLS:
-  upload_files     - Upload files to Hailer
-  download_file    - Download file from Hailer
-
-USER TOOLS:
-  search_workspace_users - Find users by name
-
-TEMPLATE TOOLS:
-  list_templates     - List marketplace templates
-  get_template       - Get template details
-  install_template   - Install template to workspace
-  publish_template   - Publish workspace as template
-
-USAGE:
-  Tools are called via the MCP protocol. The orchestrator and
-  agents use them automatically based on your requests.
-
-SEE ALSO:
-  /help agents   - How agents use these tools
-  /help commands - Slash commands available
-```

package/.claude/commands/hotfix-squad.md

@@ -1,112 +0,0 @@
----
-description: Quick bug fix with test creation, code editing, and security review
-argument-hint: "bug description"
-allowed-tools: Task, Bash
----
-# Hotfix Squad
-
-Sequential pipeline with iterative loop: Reproduce bug, apply fix, verify, review. Loops back if review fails.
-
-**Roles:**
-1. **Test Runner** - Test runner (create failing test, verify fix)
-2. **Code Editor** - Code editor (apply minimal fix)
-3. **Code Reviewer** - Code review (security audit of the fix)
-
-**Goal:** $ARGUMENTS
-
-## Protocol
-
-### Step 1: Reproduce Bug
-
-Spawn a test runner to create a test that demonstrates the bug:
-
-```
-Task(prompt="Load testing-patterns skill for writing focused tests.\n\nCreate a test that reproduces this bug: $ARGUMENTS.\n\nThe test should:\n- Fail with the current code (red)\n- Clearly demonstrate the bug\n- Be minimal and focused\n\nDo NOT fix the bug yet. Just create the failing test and report the failure.")
-```
-
-Wait for result. The test should be failing (RED state).
-
-### Step 2: Fix-Verify-Review Loop
-
-Track iteration count. Max **3 iterations** before escalating to user with remaining issues.
-
-**Set:** `iteration = 1`
-
-#### Step 2a: Apply Fix
-
-Spawn a code editor with accumulated context from all previous iterations:
-
-```
-Task(prompt="Fix the bug in: $ARGUMENTS.\n\nContext:\nTest runner created a failing test that reproduces the bug:\n[PASTE TEST CODE AND FAILURE OUTPUT]\n\n[IF iteration > 1: Include ALL accumulated feedback:\n- Test runner's latest test results (pass/fail details)\n- Code Reviewer's feedback (if review was reached in previous iteration)\n- What was already tried and why it didn't fully work\nThis prevents repeating previous mistakes.]\n\nMake the MINIMAL change needed to make the test pass (GREEN state). Do not refactor or improve surrounding code.")
-```
-
-#### Step 2b: Verify Fix
-
-Run the test to confirm the fix works:
-
-```
-Task(prompt="Load testing-patterns skill for test execution.\n\nRe-run the test for: $ARGUMENTS. Confirm it now passes with the code editor's fix. Also run any related tests to check for regressions.")
-```
-
-If tests FAIL:
-- If `iteration < 3`: increment iteration, go back to **Step 2a** with the test failure output
-- If `iteration >= 3`: escalate to user (see Step 3)
-
-#### Step 2c: Security Review
-
-Tests passed. Spawn a code reviewer to audit the fix:
-
-```
-Task(prompt="Security review of the hotfix for: $ARGUMENTS.\n\nFix applied by code editor:\n[PASTE CODE EDITOR'S CHANGES]\n\nIteration: [iteration] of 3\n\nCheck for:\n- Security vulnerabilities\n- Data validation issues\n- Authentication/authorization concerns\n- Edge cases that could be exploited\n\nReturn verdict: APPROVED or REQUEST_CHANGES with specific issues to fix.")
-```
-
-**If APPROVED:** proceed to Step 3 (report).
-
-**If REQUEST_CHANGES:**
-- If `iteration < 3`: increment iteration, go back to **Step 2a** with Code Reviewer's feedback
-- If `iteration >= 3`: escalate to user (see Step 3)
-
-### Step 3: Report
-
-```markdown
-## Hotfix Squad Complete
-
-### Loop Summary
-- Iterations: [count] of 3 max
-- Final status: FIXED / ESCALATED
-
-### Test Runner
-- Bug reproduced: YES
-- Test status before fix: FAILED
-- Test status after fix: PASSED / STILL FAILING
-- Regressions found: [yes/no]
-
-### Code Editor
-- Files modified: [list]
-- Lines changed: [count]
-- Approach: [brief description]
-
-### Code Reviewer
-- Verdict: APPROVED / REQUEST_CHANGES
-- Critical issues: [count]
-- Warnings: [count]
-- [If REQUEST_CHANGES: list remaining issues]
-
-### Next Steps
-[If FIXED:]
-- Deploy to staging
-- Monitor error logs
-
-[If ESCALATED after 3 iterations:]
-- Manual intervention needed
-- Remaining issues: [list from security reviewer/test runner]
-- Attempted fixes: [summary of what was tried]
-```
-
-## Notes
-
-- RED -> GREEN -> REVIEW loop ensures quality
-- Fix must be minimal - no refactoring
-- Max 3 iterations prevents infinite loops
-- Each iteration passes previous feedback to the code editor so it doesn't repeat mistakes
-- Security check is mandatory before reporting success

package/.claude/commands/integration-squad.md

@@ -1,82 +0,0 @@
----
-description: Build backend integrations with schema discovery and automation
-argument-hint: "integration description"
-allowed-tools: Task, Bash, Read
----
-# Integration Squad
-
-Sequential pipeline: Schema discovery, build integration, then verify with tests.
-
-**Roles:**
-1. **Data Reader** - Schema discovery (workflow fields, types, IDs)
-2. **Automation Builder** - Build integration (monolith webhooks, scheduled jobs, automations)
-3. **Test Runner** - Test runner (verify integration with tests)
-
-**Goal:** $ARGUMENTS
-
-## Protocol
-
-### Step 1: Schema Discovery
-
-Start with schema discovery to gather workflow schema and field IDs needed for the integration:
-
-```
-Task(prompt="Load sdk-ws-config-skill for understanding workspace structure.\n\nList all workflows and their fields relevant to: $ARGUMENTS. Include field IDs, types, and any existing integrations. Also identify webhook URLs and activity phase names needed.")
-```
-
-Wait for result. Save the schema output - the automation builder needs these IDs.
-
-### Step 2: Build Integration
-
-Spawn an automation builder with the schema context:
-
-```
-Task(prompt="Load hailer-monolith-automations skill for building integrations.\n\nBuild a backend integration for: $ARGUMENTS.\n\nWorkflow schema from data reader:\n[PASTE SCHEMA DISCOVERY OUTPUT - workflow IDs, field IDs, types, phases]\n\nImplement in the hailer-integration monolith. Use correct IDs for webhooks, scheduled jobs, or automations.")
-```
-
-Wait for result. If `"status": "ready_to_push"`:
-- Run the push commands provided
-
-### Step 3: Test Integration
-
-Spawn a test runner to verify the integration works:
-
-```
-Task(prompt="Load testing-patterns skill for integration testing.\n\nTest the integration for: $ARGUMENTS.\n\nTest coverage needed:\n- Webhook payload validation (if webhook)\n- Scheduled job execution (if scheduled)\n- Activity creation/phase transitions\n- Error handling and edge cases\n\nRun tests and report pass/fail counts.")
-```
-
-Wait for result. If tests fail, report which ones and offer to have the automation builder fix issues.
-
-### Step 4: Report
-
-```markdown
-## Integration Squad Complete
-
-### Data Reader (Schema Discovery)
-- Workflows found: [list]
-- Fields mapped: [count]
-- Phases identified: [list]
-- Existing integrations: [list or "none"]
-
-### Automation Builder
-- Integration type: [webhook/scheduled/automation]
-- Files created/modified: [list]
-- Endpoints: [list if webhook]
-- Commands run: [list]
-
-### Test Runner
-- Tests: X passed, X failed
-- Build: Pass/Fail
-- [Failed test details if any]
-
-### Next Steps
-- Deploy to staging environment
-- Test end-to-end in Hailer
-- Monitor for errors in production logs
-```
-
-## Notes
-
-- Data Reader must run first to get real IDs - never let automation builder guess IDs
-- Each agent in the pipeline depends on output from the previous step
-- If tests fail, the automation builder may need to revise the integration code

package/.claude/commands/janitor-squad.md

@@ -1,167 +0,0 @@
----
-description: Deep project cleanup - code, config, and data patterns
-argument-hint: [scope or blank for full project]
-allowed-tools: Task, Bash
----
-# Janitor Squad
-
-Deep project cleanup across code, configuration, and data. Discovers dead patterns and auto-fixes clear-cut removals.
-
-**Roles:**
-- **Dead Code Inspector** - Dead code detection (unused imports, unreachable code, type errors via LSP)
-- **Config Auditor** - Configuration audit (deprecated fields, broken references, unused workflows)
-- **Data Analyzer** - Data analysis (unpopulated fields, zero-activity workflows, silent insights)
-- **Code Fixer** - Auto-fix confirmed dead code and unused imports (fix loop)
-- **Code Simplifier** - Simplify complex code blocks identified during discovery (fix loop)
-
-**Target:** $ARGUMENTS (if blank, scan full project)
-
-**Changed files:** Orchestrator runs `git diff --name-only HEAD` to scope if needed.
-
-## Protocol
-
-### Step 1: Determine Scope
-
-If `$ARGUMENTS` has specific files/directories, use those as the target.
-Otherwise, scan the entire project.
-
-If `$ARGUMENTS` contains `--code-only`:
-- Skip Config Auditor and Data Analyzer
-- Only run Dead Code Inspector in discovery phase
-
-If `$ARGUMENTS` contains `--config-only`:
-- Skip Dead Code Inspector and Data Analyzer
-- Only run Config Auditor in discovery phase
-
-If `$ARGUMENTS` contains `--report-only`:
-- Skip all fix loop iterations
-- Report findings only (no auto-fixes)
-
-### Step 2: Launch Discovery Agents in Parallel
-
-**Set:** `iteration = 1`
-
-Spawn all three discovery agents simultaneously using multiple Task tool calls in a single message:
-
-**Dead Code Inspector:**
-```
-Task(prompt="Load lsp-setup skill for code inspection.\n\nInspect the project for dead code patterns: [TARGET]. Find:\n- Unused imports\n- Unreachable code blocks\n- Unused variables and functions\n- Type errors\n\nUse LSP if available, fall back to tsc/eslint. Return detailed findings grouped by file.")
-```
-
-**Config Auditor:**
-```
-Task(prompt="Audit workspace configuration for zombie patterns: [TARGET]. Check:\n- Deprecated fields in workspace/fields.ts\n- Broken field references in phases\n- Unused workflows (defined but zero activities)\n- Broken enum references\n- Orphaned configuration entries\n\nReturn issues grouped by category with severity.")
-```
-
-**Data Analyzer:**
-```
-Task(prompt="Load sdk-insight-queries skill for data analysis.\n\nAnalyze data patterns in workflows: [TARGET]. Find:\n- Fields that are defined but never populated (zero values across all activities)\n- Workflows with zero activities (created but unused)\n- Insights that return empty data (broken or obsolete)\n- Calculated fields that never execute\n- Orphaned data patterns\n\nReturn findings with field IDs, workflow IDs, and impact estimates.")
-```
-
-### Step 3: Categorize Findings
-
-Collect results from all three agents. Categorize each finding as:
-
-**AUTO-FIXABLE** (Code Fixer can safely remove):
-- Unused imports identified by Dead Code Inspector
-- Dead code blocks with no side effects
-- Workflow definitions with zero activities (from Data Analyzer + Config Auditor)
-- Deprecated enum references (from Config Auditor)
-- Never-populated fields that can be safely archived (from Data Analyzer + Config Auditor)
-
-**NEEDS-HUMAN-REVIEW** (escalate to user):
-- Architecture concerns
-- Fields that are rarely used but intentional
-- Complex logic entangled with dead code
-- Data cleanup that affects reporting history
-- Performance optimizations requiring design changes
-
-### Step 4: Auto-Fix Loop (if auto-fixable items found AND `--report-only` not set AND `iteration < 3`)
-
-If critical auto-fixable items exist, spawn two agents to clean up. If the Code Fixer's fixes introduce new issues detected by re-audit, escalate remaining issues to user after max 2 iterations.
-
-**Code Fixer:**
-```
-Task(prompt="Remove confirmed dead code and unused imports from the project:\n\n[LIST AUTO-FIXABLE ITEMS WITH FILE PATHS AND LINE NUMBERS]\n\nFrom Dead Code Inspector:\n- [Unused imports with files]\n- [Dead code blocks with files]\n\nFrom Config Auditor:\n- [Deprecated enum references with files]\n- [Unused workflow definitions with IDs]\n\nFrom Data Analyzer:\n- [Never-populated field definitions with IDs]\n\nApply minimal removals only. Do not refactor or improve beyond confirmed dead patterns.")
-```
-
-**Code Simplifier:**
-```
-Task(prompt="After dead code removal, simplify complex code blocks identified as difficult to maintain: [TARGET].\n\nFocus on:\n- Variable naming clarity\n- Function complexity reduction\n- Comment clarity around remaining logic\n- Code organization in affected files\n\nReturn refactoring suggestions with before/after examples.")
-```
-
-After both finish:
-- Increment `iteration`
-- Go back to **Step 2** to re-run all discovery agents
-- On re-discovery, tell agents to verify previous issues are resolved
-- Max 2 re-check iterations (3 total discovery runs)
-
-### Step 5: Report
-
-```markdown
-## Janitor Squad Report
-
-### Loop Summary
-- Discovery iterations: [count]
-- Auto-fixes applied: [yes/no]
-- Items removed: [count by category]
-
-### Dead Code Inspector
-- Unused imports: X (Y auto-removed)
-- Dead code blocks: X (Y auto-removed)
-- Type errors: X
-- [Remaining issues grouped by file]
-
-### Config Auditor
-- Deprecated fields: X
-- Broken references: X (Y auto-removed)
-- Unused workflows: X (Y auto-removed)
-- Orphaned config: X
-- [Remaining issues grouped by category]
-
-### Data Analyzer
-- Never-populated fields: X (Y marked for archival)
-- Zero-activity workflows: X (Y removed)
-- Silent insights: X
-- Orphaned data patterns: X
-- [Remaining items grouped by workflow]
-
-### Auto-Fixes Applied
-[If any fixes were made:]
-- [File/ID]: [what was removed/archived]
-- [File/ID]: [what was removed/archived]
-
-### Remaining Issues (Manual Action Needed)
-[Issues that require human review]
-- [Category]: [Details]
-
-### Recommended Cleanup Order
-1. [Remove these orphaned items first]
-2. [Archive these fields next]
-3. [Review these design concerns]
-
-### Summary
-[Overall health assessment]
-[Estimated time to apply remaining cleanup]
-```
-
-If any agent fails or times out, report partial results from successful agents, note which failed, and offer to re-run individually.
-
-## Options
-
-| Flag | Effect |
-|------|--------|
-| `--code-only` | Skip Config Auditor and Data Analyzer, just code cleanup via Dead Code Inspector |
-| `--config-only` | Skip Dead Code Inspector and Data Analyzer, just workspace config audit via Config Auditor |
-| `--report-only` | Report findings only, no auto-fix loop (discovery phase only) |
-| `--bg` | Run all discovery agents in background mode |
-
-If `--bg` is present, launch all discovery agents with `run_in_background: true` and tell the user they'll be notified when discovery is complete, then proceed with fix loop when they're ready.
-
-## Notes
-
-- Max 2 re-discovery iterations (3 total) to prevent infinite loops
-- Code Fixer only removes clear-cut dead code - ambiguous items go to user review
-- Data cleanup (Data Analyzer findings) may affect reports/history - always prompt user before removal
-- Each re-discovery is a full parallel sweep across all enabled agents
-- The `--report-only` flag gives discovery results without auto-fixes