@every-env/compound-plugin 2.35.0 → 2.36.1

package/.claude-plugin/marketplace.json

@@ -11,8 +11,8 @@
   "plugins": [
     {
       "name": "compound-engineering",
-      "description": "AI-powered development tools that get smarter with every use. Make each unit of engineering work easier than the last. Includes 29 specialized agents, 22 commands, and 20 skills.",
-      "version": "2.38.1",
+      "description": "AI-powered development tools that get smarter with every use. Make each unit of engineering work easier than the last. Includes 28 specialized agents and 47 skills.",
+      "version": "2.40.0",
       "author": {
         "name": "Kieran Klaassen",
         "url": "https://github.com/kieranklaassen",

package/CHANGELOG.md

@@ -7,6 +7,29 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 Release numbering now follows the repository `v*` tag line. Starting at `v2.34.0`, the root CLI package and this changelog stay on that shared version stream. Older entries below retain the previous `0.x` CLI numbering.
 
+## [2.36.1](https://github.com/EveryInc/compound-engineering-plugin/compare/v2.36.0...v2.36.1) (2026-03-13)
+
+
+### Bug Fixes
+
+* **agents:** update learnings-researcher model from haiku to inherit ([30852b7](https://github.com/EveryInc/compound-engineering-plugin/commit/30852b72937091b0a85c22b7c8c45d513ab49fd1)), closes [#249](https://github.com/EveryInc/compound-engineering-plugin/issues/249)
+
+# [2.36.0](https://github.com/EveryInc/compound-engineering-plugin/compare/v2.35.0...v2.36.0) (2026-03-11)
+
+
+### Bug Fixes
+
+* **hooks:** wrap PreToolUse handlers in try-catch to prevent parallel tool call crashes ([598222e](https://github.com/EveryInc/compound-engineering-plugin/commit/598222e11cb2206a2e3347cb5dd38cacdc3830df)), closes [#85](https://github.com/EveryInc/compound-engineering-plugin/issues/85)
+* **install:** merge config instead of overwriting on opencode target ([1db7680](https://github.com/EveryInc/compound-engineering-plugin/commit/1db76800f91fefcc1bb9c1798ef273ddd0b65f5c)), closes [#125](https://github.com/EveryInc/compound-engineering-plugin/issues/125)
+* **review:** add serial mode to prevent context limit crashes ([d96671b](https://github.com/EveryInc/compound-engineering-plugin/commit/d96671b9e9ecbe417568b2ce7f7fa4d379c2bec2)), closes [#166](https://github.com/EveryInc/compound-engineering-plugin/issues/166)
+
+
+### Features
+
+* **compound:** add context budget precheck and compact-safe mode ([c4b1358](https://github.com/EveryInc/compound-engineering-plugin/commit/c4b13584312058cb8db3ad0f25674805bbb91b2d)), closes [#198](https://github.com/EveryInc/compound-engineering-plugin/issues/198)
+* **plan:** add daily sequence number to plan filenames ([e94ca04](https://github.com/EveryInc/compound-engineering-plugin/commit/e94ca0409671efcfa2d4a8fcb2d60b79a848fd85)), closes [#135](https://github.com/EveryInc/compound-engineering-plugin/issues/135)
+* **plugin:** release v2.39.0 with community contributions ([d2ab6c0](https://github.com/EveryInc/compound-engineering-plugin/commit/d2ab6c076882a4dacaa787c0a6f3c9d555d38af0))
+
 # [2.35.0](https://github.com/EveryInc/compound-engineering-plugin/compare/v2.34.7...v2.35.0) (2026-03-10)
 
 

package/README.md

@@ -158,17 +158,20 @@ Notes:
 ## Workflow
 
 ```
-Plan → Work → Review → Compound → Repeat
+Brainstorm → Plan → Work → Review → Compound → Repeat
 ```
 
 | Command | Purpose |
 |---------|---------|
+| `/ce:brainstorm` | Explore requirements and approaches before planning |
 | `/ce:plan` | Turn feature ideas into detailed implementation plans |
 | `/ce:work` | Execute plans with worktrees and task tracking |
 | `/ce:review` | Multi-agent code review before merging |
 | `/ce:compound` | Document learnings to make future work easier |
 
-Each cycle compounds: plans inform future plans, reviews catch more issues, patterns get documented.
+The `brainstorming` skill supports `/ce:brainstorm` with collaborative dialogue to clarify requirements and compare approaches before committing to a plan.
+
+Each cycle compounds: brainstorms sharpen plans, plans inform future plans, reviews catch more issues, patterns get documented.
 
 ## Philosophy
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@every-env/compound-plugin",
-  "version": "2.35.0",
+  "version": "2.36.1",
   "type": "module",
   "private": false,
   "bin": {

package/plugins/compound-engineering/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "compound-engineering",
-  "version": "2.38.1",
-  "description": "AI-powered development tools. 28 agents, 22 commands, 20 skills, 1 MCP server for code review, research, design, and workflow automation.",
+  "version": "2.40.0",
+  "description": "AI-powered development tools. 28 agents, 47 skills, 1 MCP server for code review, research, design, and workflow automation.",
   "author": {
     "name": "Kieran Klaassen",
     "email": "kieran@every.to",

package/plugins/compound-engineering/CHANGELOG.md

@@ -5,6 +5,32 @@ All notable changes to the compound-engineering plugin will be documented in thi
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [2.39.0] - 2026-03-10
+
+### Added
+
+- **ce:compound context budget precheck** — Warns when context is constrained and offers compact-safe mode to avoid compaction mid-compound ([#235](https://github.com/EveryInc/compound-engineering-plugin/pull/235))
+- **ce:plan daily sequence numbers** — Plan filenames now include a 3-digit daily sequence number (e.g., `2026-03-10-001-feat-...`) to prevent collisions ([#238](https://github.com/EveryInc/compound-engineering-plugin/pull/238))
+- **ce:review serial mode** — Pass `--serial` flag (or auto-detects when 6+ agents configured) to run review agents sequentially, preventing context limit crashes ([#237](https://github.com/EveryInc/compound-engineering-plugin/pull/237))
+- **agent-browser inspection & debugging commands** — Added JS eval, console/errors, network, storage, device emulation, element debugging, recording/tracing, tabs, and advanced mouse commands to agent-browser skill ([#236](https://github.com/EveryInc/compound-engineering-plugin/pull/236))
+- **test-browser port detection** — Auto-detects dev server port from CLAUDE.md, package.json, or .env files; supports `--port` flag ([#233](https://github.com/EveryInc/compound-engineering-plugin/pull/233))
+- **lfg phase gating** — Added explicit GATE checks between /lfg steps to enforce plan-before-work ordering ([#231](https://github.com/EveryInc/compound-engineering-plugin/pull/231))
+
+### Fixed
+
+- **Context7 API key auth** — MCP server config now passes `CONTEXT7_API_KEY` via `x-api-key` header to avoid anonymous rate limits ([#232](https://github.com/EveryInc/compound-engineering-plugin/pull/232))
+- **CLI: MCP server merge order** — `sync` now correctly overwrites same-named MCP servers with plugin values instead of preserving stale entries
+
+### Removed
+
+- **every-style-editor agent** — Removed duplicate agent; functionality already exists as `every-style-editor` skill ([#234](https://github.com/EveryInc/compound-engineering-plugin/pull/234))
+
+### Contributors
+
+- Matt Van Horn ([@mvanhorn](https://x.com/mvanhorn)) — PRs #231–#238
+
+---
+
 ## [2.38.1] - 2026-03-01
 
 ### Fixed

package/plugins/compound-engineering/CLAUDE.md

@@ -34,15 +34,16 @@ agents/
 ├── workflow/   # Workflow automation agents
 └── docs/       # Documentation agents
 
-commands/
-├── ce/         # Core workflow commands (ce:plan, ce:review, etc.)
-├── workflows/  # Deprecated aliases for ce:* commands
-└── *.md        # Utility commands
-
 skills/
-└── *.md        # All skills at root level
+├── ce-*/          # Core workflow skills (ce:plan, ce:review, etc.)
+├── workflows-*/   # Deprecated aliases for ce:* skills
+└── */             # All other skills
 ```
 
+> **Note:** Commands were migrated to skills in v2.39.0. All former
+> `/command-name` slash commands now live under `skills/command-name/SKILL.md`
+> and work identically (Claude Code 2.1.3+ merged the two formats).
+
 ## Command Naming Convention
 
 **Workflow commands** use `ce:` prefix to unambiguously identify them as compound-engineering commands:

package/plugins/compound-engineering/README.md

@@ -6,7 +6,7 @@ AI-powered development tools that get smarter with every use. Make each unit of
 
 | Component | Count |
 |-----------|-------|
-| Agents | 29 |
+| Agents | 28 |
 | Commands | 22 |
 | Skills | 20 |
 | MCP Servers | 1 |

package/plugins/compound-engineering/agents/research/learnings-researcher.md

@@ -1,7 +1,7 @@
 ---
 name: learnings-researcher
 description: "Searches docs/solutions/ for relevant past solutions by frontmatter metadata. Use before implementing features or fixing problems to surface institutional knowledge and prevent repeated mistakes."
-model: haiku
+model: inherit
 ---
 
 <examples>

package/plugins/compound-engineering/{commands/ce/review.md → skills/ce-review/SKILL.md}

@@ -1,7 +1,7 @@
 ---
 name: ce:review
 description: Perform exhaustive code reviews using multi-agent analysis, ultra-thinking, and worktrees
-argument-hint: "[PR number, GitHub URL, branch name, or latest]"
+argument-hint: "[PR number, GitHub URL, branch name, or latest] [--serial]"
 ---
 
 # Review Command
@@ -65,17 +65,50 @@ Read `compound-engineering.local.md` in the project root. If found, use `review_
 
 If no settings file exists, invoke the `setup` skill to create one. Then read the newly created file and continue.
 
+#### Choose Execution Mode
+
+<execution_mode>
+
+Before launching review agents, check for context constraints:
+
+**If `--serial` flag is passed OR conversation is in a long session:**
+
+Run agents ONE AT A TIME in sequence. Wait for each agent to complete before starting the next. This uses less context but takes longer.
+
+**Default (parallel):**
+
+Run all agents simultaneously for speed. If you hit context limits, retry with `--serial` flag.
+
+**Auto-detect:** If more than 5 review agents are configured, automatically switch to serial mode and inform the user:
+"Running review agents in serial mode (6+ agents configured). Use --parallel to override."
+
+</execution_mode>
+
 #### Parallel Agents to review the PR:
 
 <parallel_tasks>
 
+**Parallel mode (default for ≤5 agents):**
+
 Run all configured review agents in parallel using Task tool. For each agent in the `review_agents` list:
 
 ```
 Task {agent-name}(PR content + review context from settings body)
 ```
 
-Additionally, always run these regardless of settings:
+**Serial mode (--serial flag, or auto for 6+ agents):**
+
+Run configured review agents ONE AT A TIME. For each agent in the `review_agents` list, wait for it to complete before starting the next:
+
+```
+For each agent in review_agents:
+  1. Task {agent-name}(PR content + review context)
+  2. Wait for completion
+  3. Collect findings
+  4. Proceed to next agent
+```
+
+Always run these last regardless of mode:
 - Task agent-native-reviewer(PR content) - Verify new features are agent-accessible
 - Task learnings-researcher(PR content) - Search docs/solutions/ for past issues related to this PR's modules and patterns
 

package/plugins/compound-engineering/{commands/deploy-docs.md → skills/deploy-docs/SKILL.md}

@@ -15,7 +15,6 @@ Run these checks:
 ```bash
 # Count components
 echo "Agents: $(ls plugins/compound-engineering/agents/*.md | wc -l)"
-echo "Commands: $(ls plugins/compound-engineering/commands/*.md | wc -l)"
 echo "Skills: $(ls -d plugins/compound-engineering/skills/*/ 2>/dev/null | wc -l)"
 
 # Validate JSON

package/plugins/compound-engineering/{commands/generate_command.md → skills/generate_command/SKILL.md}

@@ -7,7 +7,7 @@ disable-model-invocation: true
 
 # Create a Custom Claude Code Command
 
-Create a new slash command in `.claude/commands/` for the requested task.
+Create a new skill in `.claude/skills/` for the requested task.
 
 ## Goal
 
@@ -128,10 +128,10 @@ Implement #$ARGUMENTS following these steps:
 
 ## Creating the Command File
 
-1. **Create the file** at `.claude/commands/[name].md` (subdirectories like `workflows/` supported)
+1. **Create the directory** at `.claude/skills/[name]/SKILL.md`
 2. **Start with YAML frontmatter** (see section above)
-3. **Structure the command** using the template above
-4. **Test the command** by using it with appropriate arguments
+3. **Structure the skill** using the template above
+4. **Test the skill** by using it with appropriate arguments
 
 ## Command File Template
 

package/src/converters/claude-to-opencode.ts

@@ -202,7 +202,15 @@ function renderHookHandlers(
   const wrapped = options.requireError
     ? `    if (input?.error) {\n${statements.map((line) => `      ${line}`).join("\n")}\n    }`
     : rendered
+
+  // Wrap tool.execute.before handlers in try-catch to prevent a failing hook
+  // from crashing parallel tool call batches (causes API 400 errors).
+  // See: https://github.com/EveryInc/compound-engineering-plugin/issues/85
+  const isPreToolUse = event === "tool.execute.before"
   const note = options.note ? `    // ${options.note}\n` : ""
+  if (isPreToolUse) {
+    return `    "${event}": async (input) => {\n${note}    try {\n  ${wrapped}\n    } catch (err) {\n      console.error("[hook] ${event} error (non-fatal):", err)\n    }\n    }`
+  }
   return `    "${event}": async (input) => {\n${note}${wrapped}\n    }`
 }
 

package/src/sync/json-config.ts

@@ -19,7 +19,7 @@ export async function mergeJsonConfigAtKey(options: {
     ...existing,
     [key]: {
       ...existingEntries,
-      ...incoming,
+      ...incoming, // incoming plugin entries overwrite same-named servers
     },
   }
 

package/src/targets/opencode.ts

@@ -58,12 +58,16 @@ export async function writeOpenCodeBundle(outputRoot: string, bundle: OpenCodeBu
   const openCodePaths = resolveOpenCodePaths(outputRoot)
   await ensureDir(openCodePaths.root)
 
+  const hadExistingConfig = await pathExists(openCodePaths.configPath)
   const backupPath = await backupFile(openCodePaths.configPath)
   if (backupPath) {
     console.log(`Backed up existing config to ${backupPath}`)
   }
   const merged = await mergeOpenCodeConfig(openCodePaths.configPath, bundle.config)
   await writeJson(openCodePaths.configPath, merged)
+  if (hadExistingConfig) {
+    console.log("Merged plugin config into existing opencode.json (user settings preserved)")
+  }
 
   const agentsDir = openCodePaths.agentsDir
   for (const agent of bundle.agents) {

package/tests/converter.test.ts

@@ -132,6 +132,18 @@ describe("convertClaudeToOpenCode", () => {
     expect(hookFile!.content).toContain("// timeout: 30s")
     expect(hookFile!.content).toContain("// Prompt hook for Write|Edit")
     expect(hookFile!.content).toContain("// Agent hook for Write|Edit: security-sentinel")
+
+    // PreToolUse (tool.execute.before) handlers are wrapped in try-catch
+    // to prevent hook failures from crashing parallel tool call batches (#85)
+    const beforeIdx = hookFile!.content.indexOf('"tool.execute.before"')
+    const afterIdx = hookFile!.content.indexOf('"tool.execute.after"')
+    const beforeBlock = hookFile!.content.slice(beforeIdx, afterIdx)
+    expect(beforeBlock).toContain("try {")
+    expect(beforeBlock).toContain("} catch (err) {")
+
+    // PostToolUse (tool.execute.after) handlers are NOT wrapped in try-catch
+    const afterBlock = hookFile!.content.slice(afterIdx, hookFile!.content.indexOf('"session.created"'))
+    expect(afterBlock).not.toContain("try {")
   })
 
   test("converts MCP servers", async () => {

package/tests/opencode-writer.test.ts

@@ -3,6 +3,7 @@ import { promises as fs } from "fs"
 import path from "path"
 import os from "os"
 import { writeOpenCodeBundle } from "../src/targets/opencode"
+import { mergeJsonConfigAtKey } from "../src/sync/json-config"
 import type { OpenCodeBundle } from "../src/types/opencode"
 
 async function exists(filePath: string): Promise<boolean> {
@@ -254,3 +255,38 @@ describe("writeOpenCodeBundle", () => {
     expect(backupContent).toBe("old content\n")
   })
 })
+
+describe("mergeJsonConfigAtKey", () => {
+  test("incoming plugin entries overwrite same-named servers", async () => {
+    const tempDir = await fs.mkdtemp(path.join(os.tmpdir(), "json-merge-"))
+    const configPath = path.join(tempDir, "opencode.json")
+
+    // User has an existing MCP server config
+    const existingConfig = {
+      model: "my-model",
+      mcp: {
+        "user-server": { type: "local", command: ["uvx", "user-srv"] },
+      },
+    }
+    await fs.writeFile(configPath, JSON.stringify(existingConfig, null, 2))
+
+    // Plugin syncs its servers, overwriting same-named entries
+    await mergeJsonConfigAtKey({
+      configPath,
+      key: "mcp",
+      incoming: {
+        "plugin-server": { type: "local", command: ["uvx", "plugin-srv"] },
+        "user-server": { type: "local", command: ["uvx", "plugin-override"] },
+      },
+    })
+
+    const merged = JSON.parse(await fs.readFile(configPath, "utf8"))
+
+    // User's top-level keys preserved
+    expect(merged.model).toBe("my-model")
+    // Plugin server added
+    expect(merged.mcp["plugin-server"]).toBeDefined()
+    // Plugin server overwrites same-named existing entry
+    expect(merged.mcp["user-server"].command[1]).toBe("plugin-override")
+  })
+})

package/plans/grow-your-own-garden-plugin-architecture.md

@@ -1,102 +0,0 @@
-# Grow Your Own Garden: Adaptive Agent Ecosystem
-
-> **Issue:** https://github.com/EveryInc/compound-engineering-plugin/issues/20
-
-## The Idea
-
-Everyone grows their own garden, but we're all using the same process.
-
-Start from a **seed** (minimal core: `/plan`, `/work`, `/review`, `/compound`). Each `/compound` loop can suggest adding agents based on what you're working on—like building up a test suite to prevent regressions, but for code review expertise.
-
-## Current Problem
-
-- Monolithic plugin: 24 agents, users use ~30%
-- No personalization (same agents for Rails dev and Python dev)
-- Static collection that doesn't adapt
-
-## Proposed Solution
-
-### The Seed (Core Plugin)
-
-4 commands + minimal agents:
-
-| Component | What's Included |
-|-----------|-----------------|
-| Commands | `/plan`, `/work`, `/review`, `/compound` |
-| Review Agents | security, performance, simplicity, architecture, patterns |
-| Research Agents | best-practices, framework-docs, git-history, repo-analyst |
-| Skills | compound-docs, file-todos, git-worktree |
-| MCP Servers | playwright, context7 |
-
-### The Growth Loop
-
-After each `/compound`:
-
-```
-✅ Learning documented
-
-💡 It looks like you're using Rails.
-   Would you like to add the "DHH Rails Reviewer"?
-
-   [y] Yes  [n] No  [x] Never ask
-```
-
-Three sources of new agents:
-1. **Predefined** - "You're using Rails, add DHH reviewer?"
-2. **Dynamic** - "You're using actor model, create an expert?"
-3. **Custom** - "Want to create an agent for this pattern?"
-
-### Agent Storage
-
-```
-.claude/agents/       → Project-specific (highest priority)
-~/.claude/agents/     → User's garden
-plugin/agents/        → From installed plugins
-```
-
-## Implementation Phases
-
-### Phase 1: Split the Plugin
-- Create `agent-library/` with framework-specific agents (Rails, Python, TypeScript, Frontend)
-- Keep `compound-engineering` as core with universal agents
-- No breaking changes—existing users unaffected
-
-### Phase 2: Agent Discovery
-- `/review` discovers agents from all three locations
-- Project agents override user agents override plugin agents
-
-### Phase 3: Growth via /compound
-- Detect tech stack (Gemfile, package.json, etc.)
-- Suggest relevant agents after documenting learnings
-- Install accepted agents to `~/.claude/agents/`
-
-### Phase 4: Management
-- `/agents list` - See your garden
-- `/agents add <name>` - Add from library
-- `/agents disable <name>` - Temporarily disable
-
-## What Goes Where
-
-**Core (seed):** 11 framework-agnostic agents
-- security-sentinel, performance-oracle, code-simplicity-reviewer
-- architecture-strategist, pattern-recognition-specialist
-- 4 research agents, 2 workflow agents
-
-**Agent Library:** 10 specialized agents
-- Rails: kieran-rails, dhh-rails, data-integrity (3)
-- Python: kieran-python (1)
-- TypeScript: kieran-typescript (1)
-- Frontend: julik-races, design-iterator, design-reviewer, figma-sync (4)
-- Editorial: every-style-editor (1)
-
-## Key Constraint
-
-Claude Code doesn't support plugin dependencies. Each plugin must be independent. Users manually install what they need, or we suggest additions via `/compound`.
-
-## Acceptance Criteria
-
-- [ ] Core plugin works standalone with universal agents
-- [ ] `/compound` suggests agents based on detected tech stack
-- [ ] Users can accept/decline suggestions
-- [ ] `/agents` command for garden management
-- [ ] No breaking changes for existing users

package/plans/landing-page-launchkit-refresh.md

@@ -1,279 +0,0 @@
-# Landing Page LaunchKit Refresh
-
-## Overview
-
-Review and enhance the `/docs/index.html` landing page using LaunchKit elements and Pragmatic Technical Writing style (Hunt/Thomas, Joel Spolsky). The current implementation is strong but can be refined section-by-section.
-
-## Current State Assessment
-
-### What's Working Well
-- Specific, outcome-focused hero headline ("12 expert opinions in 30 seconds")
-- Developer-authentic copywriting (N+1 queries, CORS, SQL injection)
-- Stats section with clear metrics (23 agents, 16 commands, 11 skills, 2 MCP servers)
-- Philosophy section with concrete story (N+1 query bug)
-- Three-step installation with actual commands
-- FAQ accordion following LaunchKit patterns
-- Categorized feature sections with code examples
-
-### Missing Elements (From Best Practices Research)
-1. **Social Proof Section** - No testimonials, GitHub stars, or user metrics
-2. **Visual Demo** - No GIF/animation showing the tool in action
-3. **Arrow icons on CTAs** - 26% conversion boost from studies
-4. **Trust indicators** - Open source badge, license info
-
----
-
-## Section-by-Section Review Plan
-
-### 1. Hero Section (lines 56-78)
-
-**Current:**
-```html
-<h1>Your Code Reviews Just Got 12 Expert Opinions. In 30 Seconds.</h1>
-```
-
-**Review Checklist:**
-- [ ] Headline follows Pragmatic Writing (concrete before abstract) ✅
-- [ ] Eyebrow badge is current (Version 2.6.0) - verify
-- [ ] Description paragraph under 3 sentences ✅
-- [ ] Button group has arrow icon on primary CTA
-- [ ] "Read the Docs" secondary CTA present ✅
-
-**Potential Improvements:**
-- Add `→` arrow to "Install Plugin" button
-- Consider adding animated terminal GIF below buttons showing `/review` in action
-
-### 2. Stats Section (lines 81-104)
-
-**Current:** 4 stat cards (23 agents, 16 commands, 11 skills, 2 MCP servers)
-
-**Review Checklist:**
-- [ ] Numbers are accurate (verify against actual file counts)
-- [ ] Icons are appropriate for each stat
-- [ ] Hover effects working properly
-- [ ] Mobile layout (2x2 grid) is readable
-
-**Potential Improvements:**
-- Add "developers using" or "reviews run" metric if available
-- Consider adding subtle animation on scroll
-
-### 3. Philosophy Section (lines 107-192)
-
-**Current:** "Why Your Third Code Review Should Be Easier Than Your First" with N+1 query story
-
-**Review Checklist:**
-- [ ] Opens with concrete story (N+1 query) ✅
-- [ ] Quote block is memorable and quotable
-- [ ] Four pillars (Plan, Delegate, Assess, Codify) are clear
-- [ ] Each pillar has: tagline, description, tool tags
-- [ ] Descriptions use "you" voice ✅
-
-**Potential Improvements:**
-- Review pillar descriptions for passive voice
-- Ensure each pillar description follows PAS (Problem, Agitate, Solve) pattern
-- Check tool tags are accurate and current
-
-### 4. Agents Section (lines 195-423)
-
-**Current:** 23 agents in 5 categories (Review, Research, Design, Workflow, Docs)
-
-**Review Checklist:**
-- [ ] All 23 agents are listed (count actual files)
-- [ ] Categories are logical and scannable
-- [ ] Each card has: name, badge, description, usage code
-- [ ] Descriptions are conversational (not passive)
-- [ ] Critical badges (Security, Data) stand out
-
-**Potential Improvements:**
-- Review agent descriptions against pragmatic writing checklist
-- Ensure descriptions answer "when would I use this?"
-- Add concrete scenarios to generic descriptions
-
-### 5. Commands Section (lines 426-561)
-
-**Current:** 16 commands in 2 categories (Workflow, Utility)
-
-**Review Checklist:**
-- [ ] All 16 commands are listed (count actual files)
-- [ ] Core workflow commands are highlighted
-- [ ] Descriptions are action-oriented
-- [ ] Command names match actual implementation
-
-**Potential Improvements:**
-- Review command descriptions for passive voice
-- Lead with outcomes, not features
-- Add "saves you X minutes" framing where appropriate
-
-### 6. Skills Section (lines 564-703)
-
-**Current:** 11 skills in 3 categories (Development, Content/Workflow, Image Generation)
-
-**Review Checklist:**
-- [ ] All 11 skills are listed (count actual directories)
-- [ ] Featured skill (gemini-imagegen) is properly highlighted
-- [ ] API key requirement is clear
-- [ ] Skill invocation syntax is correct
-
-**Potential Improvements:**
-- Review skill descriptions against pragmatic writing
-- Ensure each skill answers "what problem does this solve?"
-
-### 7. MCP Servers Section (lines 706-751)
-
-**Current:** 2 MCP servers (Playwright, Context7)
-
-**Review Checklist:**
-- [ ] Tool lists are accurate
-- [ ] Descriptions explain WHY not just WHAT
-- [ ] Framework support list is current (100+)
-
-**Potential Improvements:**
-- Add concrete example of each server in action
-- Consider before/after comparison
-
-### 8. Installation Section (lines 754-798)
-
-**Current:** "Three Commands. Zero Configuration." with 3 steps
-
-**Review Checklist:**
-- [ ] Commands are accurate and work
-- [ ] Step 3 shows actual usage examples
-- [ ] Timeline visual (vertical line) renders correctly
-- [ ] Copy buttons work on code blocks
-
-**Potential Improvements:**
-- Add copy-to-clipboard functionality if missing
-- Consider adding "What you'll see" output example
-
-### 9. FAQ Section (lines 801-864)
-
-**Current:** 5 questions in accordion format
-
-**Review Checklist:**
-- [ ] Questions address real objections
-- [ ] Answers are conversational (use "you")
-- [ ] Accordion expand/collapse works
-- [ ] No passive voice in answers
-
-**Potential Improvements:**
-- Review for weasel words ("best practices suggest")
-- Ensure answers are direct and actionable
-
-### 10. CTA Section (lines 868-886)
-
-**Current:** "Install Once. Compound Forever." with Install + GitHub buttons
-
-**Review Checklist:**
-- [ ] Badge is eye-catching ("Free & Open Source")
-- [ ] Headline restates core value proposition
-- [ ] Primary CTA has arrow icon ✅
-- [ ] Trust line at bottom
-
-**Potential Improvements:**
-- Review trust line copy
-- Consider adding social proof element
-
----
-
-## NEW: Social Proof Section (To Add)
-
-**Position:** After Stats section, before Philosophy section
-
-**Components:**
-- GitHub stars counter (dynamic or static)
-- "Trusted by X developers" metric
-- 2-3 testimonial quotes (if available)
-- Company logos (if applicable)
-
-**LaunchKit Pattern:**
-```html
-<section class="social-proof-section">
-  <div class="heading centered">
-    <p class="paragraph m secondary">Trusted by developers at</p>
-  </div>
-  <div class="logo-grid">
-    <!-- Company logos or GitHub badge -->
-  </div>
-</section>
-```
-
----
-
-## Pragmatic Writing Style Checklist (Apply to ALL Copy)
-
-### The Five Laws
-1. **Concrete Before Abstract** - Story/example first, then principle
-2. **Physical Analogies** - Import metaphors readers understand
-3. **Conversational Register** - Use "you", contractions, asides
-4. **Numbered Frameworks** - Create referenceable structures
-5. **Humor as Architecture** - Mental anchors for dense content
-
-### Anti-Patterns to Find and Fix
-- [ ] "It is recommended that..." → "Do this:"
-- [ ] "Best practices suggest..." → "Here's what works:"
-- [ ] Passive voice → Active voice
-- [ ] Abstract claims → Specific examples
-- [ ] Walls of text → Scannable lists
-
-### Quality Checklist (Per Section)
-- [ ] Opens with concrete story or example?
-- [ ] Can reader skim headers and get the arc?
-- [ ] Uses "you" at least once?
-- [ ] Clear action reader can take?
-- [ ] Reads aloud like speech?
-
----
-
-## Implementation Phases
-
-### Phase 1: Copy Audit (No HTML Changes)
-1. Read through entire page
-2. Flag passive voice instances
-3. Flag abstract claims without examples
-4. Flag missing "you" voice
-5. Document improvements needed
-
-### Phase 2: Copy Rewrites
-1. Rewrite flagged sections following pragmatic style
-2. Ensure each section passes quality checklist
-3. Maintain existing HTML structure
-
-### Phase 3: Component Additions
-1. Add arrow icons to primary CTAs
-2. Add social proof section (if data available)
-3. Consider visual demo element
-
-### Phase 4: Verification
-1. Validate all counts (agents, commands, skills)
-2. Test all links and buttons
-3. Verify mobile responsiveness
-4. Check accessibility
-
----
-
-## Files to Modify
-
-| File | Changes |
-|------|---------|
-| `docs/index.html` | Copy rewrites, potential new section |
-| `docs/css/style.css` | Social proof styles (if adding) |
-
----
-
-## Success Criteria
-
-1. All copy passes Pragmatic Writing quality checklist
-2. No passive voice in any description
-3. Every feature section answers "why should I care?"
-4. Stats are accurate against actual file counts
-5. Page loads in <3 seconds
-6. Mobile layout is fully functional
-
----
-
-## References
-
-- LaunchKit Template: https://launchkit.evilmartians.io/
-- Pragmatic Writing Skill: `~/.claude/skills/pragmatic-writing-skill/SKILL.md`
-- Current Landing Page: `/Users/kieranklaassen/compound-engineering-plugin/docs/index.html`
-- Style CSS: `/Users/kieranklaassen/compound-engineering-plugin/docs/css/style.css`