Command Reference — Core Plugins
Flag and option documentation for core plugin commands (abstract, attune, conserve, imbue, sanctum).
Extended plugins: Memory Palace, Pensive, Parseltongue, Spec-Kit, Scribe, Scry, Hookify, Leyline
See also: Capabilities Reference | Skills | Agents | Hooks | Workflows
Command Syntax
/<plugin>:<command-name> [--flags] [positional-args]
Common Flag Patterns:
| Flag Pattern | Description | Example |
|---|---|---|
--verbose | Enable detailed output | /bloat-scan --verbose |
--dry-run | Preview without executing | /unbloat --dry-run |
--force | Skip confirmation prompts | /attune:init --force |
--report FILE | Output to file | /bloat-scan --report audit.md |
--level N | Set intensity/depth | /bloat-scan --level 3 |
--skip-X | Skip specific phase | /prepare-pr --skip-updates |
Abstract Plugin
/abstract:validate-plugin
Validate plugin structure against ecosystem conventions.
# Usage
/abstract:validate-plugin [plugin-name] [--strict] [--fix]
# Options
--strict Fail on warnings (not just errors)
--fix Auto-fix correctable issues
--report FILE Output validation report
# Examples
/abstract:validate-plugin sanctum
/abstract:validate-plugin --strict conserve
/abstract:validate-plugin memory-palace --fix
/abstract:create-skill
Scaffold a new skill with proper frontmatter and structure.
# Usage
/abstract:create-skill <plugin>:<skill-name> [--template basic|modular] [--category]
# Options
--template Skill template type (basic or modular with modules/)
--category Skill category for classification
--interactive Guided creation flow
# Examples
/abstract:create-skill pensive:shell-review --template modular
/abstract:create-skill imbue:new-methodology --category workflow-methodology
/abstract:create-command
Scaffold a new command with hooks and documentation.
# Usage
/abstract:create-command <plugin>:<command-name> [--hooks] [--extends]
# Options
--hooks Include lifecycle hook templates
--extends Base command or skill to extend
--aliases Comma-separated command aliases
# Examples
/abstract:create-command sanctum:new-workflow --hooks
/abstract:create-command conserve:deep-clean --extends "conserve:bloat-scan"
/abstract:create-hook
Scaffold a new hook with security-first patterns.
# Usage
/abstract:create-hook <plugin>:<hook-name> [--type] [--lang]
# Options
--type Hook event type (PreToolUse|PostToolUse|SessionStart|Stop|UserPromptSubmit)
--lang Implementation language (bash|python)
--matcher Tool matcher pattern
# Examples
/abstract:create-hook memory-palace:cache-check --type PreToolUse --lang python
/abstract:create-hook sanctum:commit-validator --type PreToolUse --matcher "Bash"
/abstract:analyze-skill
Analyze skill complexity and optimization opportunities.
# Usage
/abstract:analyze-skill <plugin>:<skill-name> [--metrics] [--suggest]
# Options
--metrics Show detailed token/complexity metrics
--suggest Generate optimization suggestions
--compare Compare against skill baselines
# Examples
/abstract:analyze-skill imbue:proof-of-work --metrics
/abstract:analyze-skill sanctum:pr-prep --suggest
/abstract:make-dogfood
Update Makefile demonstration targets to reflect current features.
# Usage
/abstract:make-dogfood [--check] [--update]
# Options
--check Verify Makefile is current (exit 1 if stale)
--update Apply updates to Makefile
--dry-run Show what would change
# Examples
/abstract:make-dogfood --check
/abstract:make-dogfood --update
/abstract:skills-eval
Evaluate skill quality across the ecosystem.
# Usage
/abstract:skills-eval [--plugin PLUGIN] [--threshold SCORE]
# Options
--plugin Limit to specific plugin
--threshold Minimum quality score (default: 70)
--output Output format (table|json|markdown)
# Examples
/abstract:skills-eval --plugin sanctum
/abstract:skills-eval --threshold 80 --output markdown
/abstract:hooks-eval
Evaluate hook security and performance.
# Usage
/abstract:hooks-eval [--plugin PLUGIN] [--security]
# Options
--plugin Limit to specific plugin
--security Focus on security patterns
--perf Focus on performance impact
# Examples
/abstract:hooks-eval --security
/abstract:hooks-eval --plugin memory-palace --perf
/abstract:evaluate-skill
Evaluate skill execution quality.
# Usage
/abstract:evaluate-skill <plugin>:<skill-name> [--metrics] [--suggestions]
# Options
--metrics Show detailed execution metrics
--suggestions Generate improvement suggestions
--compare Compare against baseline metrics
# Examples
/abstract:evaluate-skill imbue:proof-of-work --metrics
/abstract:evaluate-skill sanctum:pr-prep --suggestions
Attune Plugin
/attune:init
Initialize project with complete development infrastructure.
# Usage
/attune:init [--lang LANGUAGE] [--name NAME] [--author AUTHOR]
# Options
--lang LANGUAGE Project language: python|rust|typescript|go
--name NAME Project name (default: directory name)
--author AUTHOR Author name
--email EMAIL Author email
--python-version VER Python version (default: 3.10)
--description TEXT Project description
--path PATH Project path (default: .)
--force Overwrite existing files without prompting
--no-git Skip git initialization
# Examples
/attune:init --lang python --name my-cli
/attune:init --lang rust --author "Your Name" --force
/attune:brainstorm
Brainstorm project ideas using Socratic questioning.
# Usage
/attune:brainstorm [TOPIC] [--output FILE]
# Options
--output FILE Save brainstorm results to file
--rounds N Number of question rounds (default: 5)
--focus AREA Focus area: features|architecture|ux|technical
# Examples
/attune:brainstorm "CLI tool for data processing"
/attune:brainstorm --focus architecture --rounds 3
/attune:blueprint
Plan architecture and break down tasks.
# Usage
/attune:blueprint [--from BRAINSTORM] [--output FILE]
# Options
--from FILE Use brainstorm results as input
--output FILE Save plan to file
--depth LEVEL Planning depth: high|detailed|exhaustive
--include Include specific aspects: tests|ci|docs
# Examples
/attune:blueprint --from brainstorm.md --depth detailed
/attune:blueprint --include tests,ci
/attune:specify
Create detailed specifications from brainstorm or plan.
# Usage
/attune:specify [--from FILE] [--type TYPE]
# Options
--from FILE Input file (brainstorm or plan)
--type TYPE Spec type: technical|functional|api|data-model
--output DIR Output directory for specs
# Examples
/attune:specify --from plan.md --type technical
/attune:specify --type api --output .specify/
/attune:execute
Execute implementation tasks systematically.
# Usage
/attune:execute [--plan FILE] [--phase PHASE] [--task ID]
# Options
--plan FILE Task plan file (default: .specify/tasks.md)
--phase PHASE Execute specific phase: setup|tests|core|integration|polish
--task ID Execute specific task by ID
--parallel Enable parallel execution where marked [P]
--continue Resume from last checkpoint
# Examples
/attune:execute --plan tasks.md --phase setup
/attune:execute --task T1.2 --parallel
/attune:validate
Validate project structure against best practices.
# Usage
/attune:validate [--strict] [--fix]
# Options
--strict Fail on warnings
--fix Auto-fix correctable issues
--config Path to custom validation config
# Examples
/attune:validate --strict
/attune:validate --fix
/attune:upgrade-project
Add or update configurations in existing project.
# Usage
/attune:upgrade-project [--component COMPONENT] [--force]
# Options
--component Specific component: makefile|precommit|workflows|gitignore
--force Overwrite existing without prompting
--diff Show diff before applying
# Examples
/attune:upgrade-project --component makefile
/attune:upgrade-project --component workflows --force
Conserve Plugin
/conserve:bloat-scan
Progressive bloat detection for dead code and duplication.
# Usage
/bloat-scan [--level 1|2|3] [--focus TYPE] [--report FILE] [--dry-run]
# Options
--level 1|2|3 Scan tier: 1=quick, 2=targeted, 3=deep audit
--focus TYPE Focus area: code|docs|deps|all (default: all)
--report FILE Save report to file
--dry-run Preview findings without taking action
--exclude PATTERN Additional exclude patterns
# Scan Tiers
# Tier 1 (2-5 min): Large files, stale files, commented code, old TODOs
# Tier 2 (10-20 min): Dead code, duplicate patterns, import bloat
# Tier 3 (30-60 min): All above + cyclomatic complexity, dependency graphs
# Examples
/bloat-scan # Quick Tier 1 scan
/bloat-scan --level 2 --focus code # Targeted code analysis
/bloat-scan --level 3 --report Q1-audit.md # Deep audit with report
/conserve:unbloat
Safe bloat remediation with interactive approval.
# Usage
/unbloat [--approve LEVEL] [--dry-run] [--backup]
# Options
--approve LEVEL Auto-approve level: high|medium|low|all
--dry-run Show what would be removed
--backup Create backup branch before changes
--interactive Prompt for each item (default)
# Examples
/unbloat --dry-run # Preview all removals
/unbloat --approve high --backup # Auto-approve high priority, backup first
/unbloat --interactive # Approve each item manually
/conserve:optimize-context
Optimize context window usage.
# Usage
/optimize-context [--target PERCENT] [--scope PATH]
# Options
--target PERCENT Target context utilization (default: 50%)
--scope PATH Limit to specific directory
--suggest Only show suggestions, don't apply
--aggressive Apply all optimizations
# Examples
/optimize-context --target 40%
/optimize-context --scope plugins/sanctum/ --suggest
/conserve:analyze-growth
Analyze skill growth patterns.
# Usage
/analyze-growth [--plugin PLUGIN] [--days N] [--trend]
# Options
--plugin PLUGIN Limit to specific plugin
--days N Analysis period (default: 30)
--trend Show growth trend predictions
--alert Alert if growth exceeds threshold
# Examples
/analyze-growth --plugin conserve --days 60
/analyze-growth --trend --alert
Imbue Plugin
/imbue:catchup
Quick context recovery after session restart.
# Usage
/catchup [--depth LEVEL] [--focus AREA]
# Options
--depth LEVEL Recovery depth: shallow|standard|deep (default: standard)
--focus AREA Focus on: git|docs|issues|all
--since DATE Catch up from specific date
# Examples
/catchup # Standard recovery
/catchup --depth deep # Full context recovery
/catchup --focus git --since "3 days ago"
/imbue:feature-review
Feature prioritization and gap analysis.
# Usage
/feature-review [--scope BRANCH] [--against BASELINE]
# Options
--scope BRANCH Review specific branch
--against BASELINE Compare against baseline (main|tag|commit)
--gaps Focus on gap analysis
--priorities Generate priority rankings
# Examples
/feature-review --scope feature/new-api
/feature-review --gaps --against main
/imbue:structured-review
Structured review workflow with methodology options.
# Usage
/structured-review PATH [--methodology METHOD]
# Options
--methodology METHOD Review methodology: evidence-based|checklist|formal
--todos Generate TodoWrite items
--summary Include executive summary
# Examples
/structured-review plugins/sanctum/ --methodology evidence-based
/structured-review . --todos --summary
Sanctum Plugin
/sanctum:prepare-pr (alias: /pr)
Complete PR preparation workflow.
# Usage
/prepare-pr [--no-code-review] [--reviewer-scope SCOPE] [--skip-updates] [FILE]
/pr [options...] # Alias
# Options
--no-code-review Skip automated code review (faster)
--reviewer-scope SCOPE Review strictness: strict|standard|lenient
--skip-updates Skip documentation/test updates (Phase 0)
FILE Output file for PR description (default: pr_description.md)
# Reviewer Scope Levels
# strict - All suggestions must be addressed
# standard - Critical issues must be fixed, suggestions are recommendations
# lenient - Focus on blocking issues only
# Examples
/prepare-pr # Full workflow
/pr # Alias for full workflow
/prepare-pr --skip-updates # Skip Phase 0 updates
/prepare-pr --no-code-review # Skip code review
/prepare-pr --reviewer-scope strict # Strict review for critical changes
/prepare-pr --skip-updates --no-code-review # Fastest (legacy behavior)
/sanctum:commit-msg
Generate commit message.
# Usage
/commit-msg [--type TYPE] [--scope SCOPE]
# Options
--type TYPE Force commit type: feat|fix|docs|refactor|test|chore
--scope SCOPE Force commit scope
--breaking Include breaking change footer
--issue N Reference issue number
# Examples
/commit-msg
/commit-msg --type feat --scope api
/commit-msg --breaking --issue 42
/sanctum:do-issue
Fix GitHub issues.
# Usage
/do-issue ISSUE_NUMBER [--branch NAME]
# Options
--branch NAME Branch name (default: issue-N)
--auto-merge Attempt auto-merge after PR
--draft Create draft PR
# Examples
/do-issue 42
/do-issue 123 --branch fix/auth-bug
/do-issue 99 --draft
/sanctum:fix-pr
Address PR review comments.
# Usage
/fix-pr [PR_NUMBER] [--auto-resolve]
# Options
PR_NUMBER PR number (default: current branch's PR)
--auto-resolve Auto-resolve addressed comments
--batch Address all comments in batch
--interactive Address one comment at a time
# Examples
/fix-pr 42
/fix-pr --auto-resolve
/fix-pr 42 --batch
/sanctum:fix-workflow
Workflow retrospective with automatic improvement context.
# Usage
/fix-workflow [WORKFLOW_NAME] [--context]
# Options
WORKFLOW_NAME Specific workflow to analyze
--context Gather improvement context automatically
--lessons Generate lessons learned
--improvements Suggest workflow improvements
# Examples
/fix-workflow pr-review --context
/fix-workflow --lessons --improvements
/sanctum:pr-review
Enhanced PR review.
# Usage
/pr-review [PR_NUMBER] [--thorough]
# Options
PR_NUMBER PR to review (default: current)
--thorough Deep review with all checks
--quick Fast review of critical issues only
--security Security-focused review
# Examples
/pr-review 42
/pr-review --thorough
/pr-review --quick --security
/sanctum:update-docs
Update project documentation.
# Usage
/update-docs [--scope SCOPE] [--check]
# Options
--scope SCOPE Scope: all|api|readme|guides
--check Check only, don't modify
--sync Sync with code changes
# Examples
/update-docs
/update-docs --scope api
/update-docs --check
/sanctum:update-readme
Modernize README.
# Usage
/update-readme [--badges] [--toc]
# Options
--badges Update/add badges
--toc Update table of contents
--examples Update code examples
--full Full README refresh
# Examples
/update-readme
/update-readme --badges --toc
/update-readme --full
/sanctum:update-tests
Maintain tests.
# Usage
/update-tests [PATH] [--coverage]
# Options
PATH Test path to update
--coverage Ensure coverage targets
--missing Add missing tests
--modernize Update to modern patterns
# Examples
/update-tests tests/
/update-tests --missing --coverage
/sanctum:update-version
Bump versions.
# Usage
/update-version [VERSION] [--type TYPE]
# Options
VERSION Explicit version (e.g., 1.2.3)
--type TYPE Bump type: major|minor|patch|prerelease
--tag Create git tag
--push Push tag to remote
# Examples
/update-version 2.0.0
/update-version --type minor --tag
/update-version --type patch --tag --push
/sanctum:update-dependencies
Update project dependencies.
# Usage
/update-dependencies [--type TYPE] [--dry-run]
# Options
--type TYPE Dependency type: all|prod|dev|security
--dry-run Preview updates without applying
--major Include major version updates
--security Security updates only
# Examples
/update-dependencies
/update-dependencies --dry-run
/update-dependencies --type security
/update-dependencies --major
/sanctum:git-catchup
Git repository catchup.
# Usage
/git-catchup [--since DATE] [--author AUTHOR]
# Options
--since DATE Start date for catchup
--author AUTHOR Filter by author
--branch BRANCH Specific branch
--format FORMAT Output format: summary|detailed|log
# Examples
/git-catchup --since "1 week ago"
/git-catchup --author "user@example.com"
/sanctum:create-tag
Create git tags for releases.
# Usage
/create-tag VERSION [--message MSG] [--sign]
# Options
VERSION Tag version (e.g., v1.0.0)
--message MSG Tag message
--sign Create signed tag
--push Push tag to remote
# Examples
/create-tag v1.0.0
/create-tag v1.0.0 --message "Release 1.0.0" --sign --push
Extended plugins: Memory Palace, Pensive, Parseltongue, Spec-Kit, Scribe, Scry, Hookify, Leyline