fix: pedagogical review - add expected output, CLAUDE.md, fix consistency

Address findings from pedagogical review simulating a non-expert user:

- Add CLAUDE.md to project root (was referenced but missing)
- Fix README score from 12/9/1 to 13/8/1 (match feature-map.md)
- Add Expected Output sections to examples 01, 02, 05, 09, 10
- Create pipeline-output/ and briefings/ directories
- Add example ordering guidance in README
- Add plan requirements for examples 11/13 in prerequisites
- Add skill frontmatter explanation in GETTING-STARTED.md
- Explain Cowork/Dispatch with links in cowork-integration
- Expand .gitignore with node_modules and generated output files
- Add model override hints in agent frontmatter comments

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

examples/01-agent-runtime/prompt.md

@@ -39,6 +39,35 @@ confirmation between steps unless it hits something ambiguous.
 
 ---
 
+## Expected Output
+
+After 30-60 seconds, you should see a new file `research-output.md` in the
+project root. It will look something like this (content varies by month):
+
+```markdown
+# AI Frameworks Released This Month
+
+## 1. ExampleFramework
+- **Released:** March 12, 2026
+- **GitHub:** https://github.com/example/framework (4,200 stars)
+- **What it solves:** Simplifies multi-agent orchestration for Python developers.
+- **Verdict:** Worth watching. Growing fast with strong community momentum.
+
+## 2. ...
+
+## Verdict
+1. ExampleFramework - most practical for production use
+2. ...
+```
+
+**How you know it worked:**
+- A file called `research-output.md` exists in the project root
+- It contains 3 frameworks with star counts and URLs
+- It ends with a ranked Verdict section
+- You saw WebSearch and WebFetch tool calls streaming in the terminal
+
+---
+
 ## Why This Matters
 
 This is the agent loop in action: plan, execute, observe, repeat. The same

examples/02-shell-and-files/prompt.md

@@ -38,6 +38,34 @@ Claude Code will:
 
 ---
 
+## Expected Output
+
+You should see Claude use 5 tool calls in sequence. The final output
+includes a directory tree and file contents:
+
+```
+project-scaffold/package.json
+project-scaffold/README.md
+project-scaffold/.gitignore
+```
+
+And the package.json contents:
+
+```json
+{
+  "name": "my-project",
+  "version": "0.1.0",
+  "description": "A scaffolded project created by Claude Code"
+}
+```
+
+**How you know it worked:**
+- A `project-scaffold/` directory exists with 3 files inside
+- You saw Bash (mkdir), Write (x3), Bash (find), and Read tool calls
+- No permission prompts appeared for these safe operations (they are pre-approved in settings.json)
+
+---
+
 ## Why This Matters
 
 Shell execution and file I/O are the foundation of every automation. Claude Code

examples/05-memory-system/prompt.md

@@ -48,6 +48,35 @@ Claude Code will:
 
 ---
 
+## Expected Output
+
+Claude creates `memory/project-notes.md` and then explains how memory works.
+The file will look like:
+
+```markdown
+# March 26, 2026
+
+- This is a companion repo comparing OpenClaw and Claude Code capabilities
+- It contains 13 examples, agents, skills, hooks, and documentation
+- The project maps 22 OpenClaw features to Claude Code equivalents
+
+Memory system demonstrated successfully.
+```
+
+Claude will then explain something like:
+
+> "This file is inside the `memory/` directory, which is referenced in
+> CLAUDE.md. Because Claude Code loads CLAUDE.md at every session start,
+> and CLAUDE.md mentions the memory directory, files here persist across
+> sessions without any extra setup."
+
+**How you know it worked:**
+- `memory/project-notes.md` exists and contains today's date
+- The three bullets accurately summarize what CLAUDE.md says
+- Claude's explanation mentions the CLAUDE.md hierarchy
+
+---
+
 ## Why This Matters
 
 OpenClaw uses SQLite-vec for semantic memory search across sessions. Claude Code

examples/09-security-hooks/prompt.md

@@ -51,6 +51,36 @@ Each entry has the format: `[timestamp] TOOL: bash | STATUS: blocked | CMD: rm -
 
 ---
 
+## Expected Output
+
+Claude will first explain what it expects the hook to do, then attempt the
+command. You should see something like:
+
+```
+I expect the PreToolUse hook (hooks/pre-tool-use.sh) to intercept this
+command because it matches the "rm -rf" pattern in the blocked list...
+
+[Claude attempts: rm -rf /tmp/test-deletion-target]
+[Hook blocks the command]
+
+The command was blocked by the PreToolUse hook. The hook matched "rm -rf"
+in the command string and returned exit code 2 with a block decision.
+```
+
+The audit log (`hooks/audit.log`) will contain an entry like:
+
+```
+[2026-03-26T10:15:23] TOOL: bash | STATUS: blocked | CMD: rm -rf /tmp/test-deletion-target
+```
+
+**How you know it worked:**
+- The `rm -rf` command was NOT executed (nothing was deleted)
+- Claude reported the hook blocked it
+- `hooks/audit.log` exists and has at least one entry
+- The entry shows STATUS: blocked
+
+---
+
 ## Architecture Difference from OpenClaw
 
 OpenClaw sandboxes via Docker: the agent runs inside a container that limits

examples/10-full-pipeline/prompt.md

@@ -62,6 +62,45 @@ typically completes in 2-4 minutes depending on web fetch latency.
 
 ---
 
+## Expected Output
+
+The pipeline takes 2-4 minutes. You will see agent invocations streaming:
+
+1. **Researcher agent** runs WebSearch and WebFetch (30-60 seconds)
+2. **Writer agent** produces a ~400-word draft (15-30 seconds)
+3. **Reviewer agent** critiques the draft (15-30 seconds)
+4. **Revision** if needed (15-30 seconds)
+5. **File writes** to pipeline-output/ and memory/
+
+When complete, two new files exist:
+
+`pipeline-output/permission-modes.md` (the article):
+```markdown
+# How Claude Code Handles Permission Modes
+
+Claude Code provides three permission modes that control how much
+autonomy the agent has...
+[~400 words with accurate technical content]
+```
+
+`memory/pipeline-log.md` (the execution log):
+```markdown
+## Pipeline Run - March 26, 2026
+
+- **Topic:** How Claude Code handles permission modes
+- **Word count:** 412
+- **Issues:** None
+- **Duration:** ~3 minutes
+```
+
+**How you know it worked:**
+- Both files exist in the expected directories
+- The article is accurate (check against `security/permission-modes-explained.md`)
+- The pipeline log has today's date and a word count
+- You saw three distinct agent invocations in the terminal
+
+---
+
 ## Why This Matters
 
 This is what Claude Code looks like as an actual agent platform, not a