feat: advanced hooks, LLM gates, and 7 new skills for v3.2.0

agents/context-engineer.md

@@ -2,6 +2,7 @@
 name: context-engineer
 description: Analyzes and optimizes context window usage across sessions. Use when context feels bloated, sessions run slow, or approaching compaction limits.
 tools: ["Read", "Glob", "Grep", "Bash"]
+omitClaudeMd: true
 ---
 
 # Context Engineer

agents/cost-analyst.md

@@ -0,0 +1,61 @@
+---
+name: cost-analyst
+description: Analyze session token usage and cost patterns. Identify expensive operations and recommend optimizations. Use to understand and reduce session costs.
+tools: ["Read", "Glob", "Grep", "Bash"]
+omitClaudeMd: true
+---
+
+# Cost Analyst
+
+Analyze token usage patterns and recommend cost optimizations.
+
+## Workflow
+
+1. Check current session token usage
+2. Identify the most expensive operations
+3. Analyze cache hit rates
+4. Recommend specific optimizations
+
+## Analysis Areas
+
+### Token Consumption by Category
+- File reads (large files without offset/limit)
+- Grep/search results (broad patterns returning many results)
+- Tool result overhead (MCP tools with verbose output)
+- System prompt size (CLAUDE.md + skills + MCP tool descriptions)
+- Agent spawning (each agent gets fresh context)
+
+### Cache Optimization
+- Stable system prompts improve cache hit rate
+- Changing CLAUDE.md mid-session breaks cache
+- Fork subagents share prompt cache (byte-identical)
+- Reusing agents via SendMessage saves context creation cost
+
+### Model Selection Impact
+- Opus: highest quality, highest cost
+- Sonnet: good balance for most tasks
+- Haiku: fast and cheap for simple lookups
+
+## Recommendations Template
+
+```text
+COST ANALYSIS
+
+Top cost drivers:
+  1. [operation] -- ~[N]K tokens
+  2. [operation] -- ~[N]K tokens
+
+Optimization opportunities:
+  - [specific change] -- saves ~[N]K tokens/session
+  - [specific change] -- saves ~$X.XX/session
+
+Current efficiency: [score]/100
+Projected savings: $X.XX/session
+```
+
+## Rules
+
+- Focus on actionable optimizations, not just reporting
+- Never sacrifice code quality for cost savings
+- Prioritize by impact (token savings x frequency)
+- Consider both input and output token costs

agents/permission-analyst.md

@@ -0,0 +1,67 @@
+---
+name: permission-analyst
+description: Analyze permission denial patterns and generate optimized alwaysAllow/alwaysDeny rules. Use when permission prompts slow down workflow.
+tools: ["Read", "Glob", "Grep", "Bash"]
+omitClaudeMd: true
+---
+
+# Permission Analyst
+
+Analyze permission patterns and recommend rule optimizations.
+
+## Workflow
+
+1. Read current permission settings from `.claude/settings.json` and `~/.claude/settings.json`
+2. Check session logs for permission approval/denial patterns
+3. Categorize operations by risk level (safe/medium/dangerous)
+4. Generate optimized rules
+
+## Risk Categories
+
+### Safe (auto-approve candidates)
+- All read-only tools: Read, Glob, Grep
+- Read-only git: `git status`, `git diff*`, `git log*`, `git branch`
+- Test/lint: `npm test*`, `npm run lint*`, `npm run typecheck*`
+- Python: `pytest*`, `ruff*`, `mypy*`
+- Rust: `cargo test*`, `cargo check*`, `cargo clippy*`
+- Go: `go test*`, `go vet*`
+
+### Medium (approve with awareness)
+- Edit, Write -- file modifications
+- `git add*` -- staging
+- `git commit*` -- committing
+- `npm install*` -- dependency changes
+
+### Dangerous (never auto-approve)
+- `git push --force*`, `git reset --hard*`
+- `rm -rf*`, `rm -r*` on non-temp dirs
+- `DROP TABLE`, `DELETE FROM` without WHERE
+- Any `--no-verify` flag
+
+## Output
+
+```text
+PERMISSION ANALYSIS
+
+Current rules: [X] allow, [Y] deny
+
+Session patterns:
+  Approved [N] times: [tool/pattern]
+  Denied [N] times: [tool/pattern]
+
+Recommended additions:
+  alwaysAllow:
+    + [rule] -- approved [N]x, [risk level]
+
+  alwaysDeny:
+    + [rule] -- [reason]
+
+Estimated prompts saved: ~[N] per session
+```
+
+## Rules
+
+- Never recommend auto-approving destructive operations
+- Present all recommendations for user approval
+- Include risk assessment for each recommendation
+- Read-only operations are always safe to auto-approve

agents/planner.md

@@ -2,6 +2,7 @@
 name: planner
 description: Break down complex tasks into implementation plans before writing code. Use when task touches >5 files, requires architecture decisions, or has unclear requirements.
 tools: ["Read", "Glob", "Grep"]
+omitClaudeMd: true
 ---
 
 # Planner

agents/reviewer.md

@@ -2,6 +2,7 @@
 name: reviewer
 description: Code review specialist that verifies every finding against actual code before reporting. Use before committing, for PR reviews, or after major changes.
 tools: ["Read", "Glob", "Grep", "Bash"]
+omitClaudeMd: true
 ---
 
 # Reviewer

agents/scout.md

@@ -4,6 +4,7 @@ description: Confidence-gated exploration that assesses readiness before impleme
 tools: ["Read", "Glob", "Grep", "Bash"]
 background: true
 isolation: worktree
+omitClaudeMd: true
 ---
 
 # Scout - Confidence-Gated Exploration

commands/auto-setup.md

@@ -0,0 +1,29 @@
+---
+description: Auto-detect project type and configure quality gates, permissions, and hooks for a new codebase
+---
+
+# /auto-setup - Project Configuration
+
+Automatically detect your project type and set up pro-workflow quality gates.
+
+## Quick Start
+
+Run in any project to:
+1. Detect project type (Node.js, Python, Rust, Go)
+2. Configure lint/typecheck/test commands
+3. Set up safe permission rules
+4. Verify tooling is installed
+
+## Supported Project Types
+
+- **Node.js/TypeScript** — npm/pnpm/yarn, ESLint, TypeScript, Jest/Vitest
+- **Python** — pip/poetry, ruff/flake8, mypy, pytest
+- **Rust** — cargo, clippy, cargo test
+- **Go** — go vet, golangci-lint, go test
+- **Mixed/Monorepo** — detects multiple types
+
+## Usage
+
+```
+/auto-setup
+```

commands/compact-guard.md

@@ -0,0 +1,35 @@
+---
+description: Smart context compaction with state preservation — saves critical state before compact and restores after
+---
+
+# /compact-guard - Protected Compaction
+
+Protect your working context through compaction.
+
+## Quick Start
+
+Run before `/compact` to:
+1. Save your current task state
+2. Note which files you're editing (max 5 survive compaction)
+3. Record decisions made this session
+4. Compact safely
+5. Restore critical context after
+
+## When To Use
+
+- Before manual `/compact`
+- When you see auto-compact warnings
+- At natural task boundaries
+
+## Key Insight
+
+Claude Code only restores **5 files** after compaction, with **5K tokens per file** and **50K total budget**. Plan accordingly:
+- Prioritize the file you're actively editing
+- Move exploration results to subagents
+- Keep notes in a scratch file
+
+## Usage
+
+```
+/compact-guard
+```

commands/cost-tracker.md

@@ -0,0 +1,30 @@
+---
+description: Track session costs, understand token spend, and get optimization tips
+---
+
+# /cost-tracker - Cost Awareness
+
+Understand and optimize your session costs.
+
+## Quick Start
+
+Run to see:
+- Current session cost estimate
+- Top cost drivers
+- Optimization suggestions
+- Budget guidance for your task type
+
+## Cost Benchmarks
+
+| Task Type | Typical | Budget Alert |
+|-----------|---------|-------------|
+| Bug fix | $0.10-0.50 | $1.00 |
+| Small feature | $0.50-2.00 | $3.00 |
+| Large feature | $2.00-8.00 | $10.00 |
+| Refactor | $1.00-5.00 | $7.00 |
+
+## Usage
+
+```
+/cost-tracker
+```

commands/mcp-audit.md

@@ -0,0 +1,31 @@
+---
+description: Audit MCP servers for token overhead, redundancy, and usage — recommend servers to disable for faster sessions
+---
+
+# /mcp-audit - MCP Server Optimization
+
+Audit your MCP servers and reduce token overhead.
+
+## Quick Start
+
+Run to see:
+- Active MCP servers and their tool counts
+- Estimated token overhead per request
+- Servers you haven't used recently
+- Recommendations for disabling/keeping
+
+## Key Insight
+
+Every MCP server adds ALL its tool descriptions to EVERY API request. A server with 20 tools adds ~2K-4K tokens per request whether you use it or not.
+
+## Thresholds
+
+- Servers: <10 ideal, >15 reduce
+- Total tools: <80 ideal, >120 reduce
+- Per server: <15 ok, >30 split or disable
+
+## Usage
+
+```
+/mcp-audit
+```

commands/permission-tuner.md

@@ -0,0 +1,32 @@
+---
+description: Analyze permission denial patterns and generate optimized allow/deny rules to reduce prompt fatigue
+---
+
+# /permission-tuner - Permission Optimization
+
+Analyze your permission patterns and generate rules to reduce prompt fatigue.
+
+## Quick Start
+
+Run this command to:
+1. Scan permission denial history
+2. Identify safe patterns for auto-approval
+3. Generate optimized rules
+4. Present for your approval
+
+## What It Does
+
+Reads your session's permission patterns and categorizes them:
+
+- **Safe to auto-approve**: Read-only operations (Read, Glob, Grep, git status/diff/log)
+- **Consider auto-approving**: Frequently approved operations (Edit, npm test)
+- **Keep asking**: Operations that need review (git commit, npm install)
+- **Auto-deny**: Dangerous operations (rm -rf, git push --force)
+
+## Usage
+
+```
+/permission-tuner
+```
+
+After running, review the suggested rules and apply the ones you want.