Skip to content

Commit e0e062f

Browse files
rysweetclaude
rysweet
and
claude
authored
docs(readme): fix Features section placement and expand feature list (#1635)
Move Features section to after Agents Reference as requested. Expand feature list from 4 to 24+ features with documentation links: **User-requested additions:** - Power-Steering - AI-powered session guidance - Auto Mode - Autonomous agentic loops - Lock Mode - Continuous work mode **Additional features documented:** - Document-Driven Development (DDD) - Fault-Tolerant Workflows (N-version, Debate, Cascade) - Security Analysis (XPIA) - Neo4j Memory System - Investigation Workflow - Skills System (54+ skills) - Fix Workflow, Reflection & Improvement - Socratic Questioning, Expert Panel - Knowledge Builder, Transcripts Management - Modular Build, Pre-commit/CI Diagnostics - Worktree Management, Session Logs - Customization System **Structural changes:** - Moved Features from lines 100-205 to after Agents Reference (line 259) - Moved Statusline under Features as subsection - Updated Table of Contents to reflect new order - Net: +160 insertions, -114 deletions Fixes #1634 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude <noreply@anthropic.com>
1 parent 23055a3 commit e0e062f

File tree

1 file changed

+160
-114
lines changed

1 file changed

+160
-114
lines changed

‎README.md‎

Lines changed: 160 additions & 114 deletions
Original file line numberDiff line numberDiff line change
@@ -15,11 +15,6 @@ Launches Claude Code with preconfigured agents. No installation needed.
1515
- [Prerequisites](#prerequisites)
1616
- [Basic Usage](#basic-usage)
1717
- [Create Alias for Easy Access](#create-alias-for-easy-access)
18-
- [Features](#features)
19-
- [Workflow Orchestration by Default](#workflow-orchestration-by-default)
20-
- [Goal-Seeking Agent Generator](#goal-seeking-agent-generator)
21-
- [Profile Management](#profile-management)
22-
- [GitHub Pages Documentation Generation](#github-pages-documentation-generation)
2318
- [Core Concepts](#core-concepts)
2419
- [Workflow](#workflow)
2520
- [Philosophy](#philosophy)
@@ -32,7 +27,13 @@ Launches Claude Code with preconfigured agents. No installation needed.
3227
- [Agents Reference](#agents-reference)
3328
- [Core Agents](#core-agents-6)
3429
- [Specialized Agents](#specialized-agents-23)
35-
- [Statusline](#statusline)
30+
- [Features](#features)
31+
- [Workflow Orchestration by Default](#workflow-orchestration-by-default)
32+
- [Goal-Seeking Agent Generator](#goal-seeking-agent-generator)
33+
- [Profile Management](#profile-management)
34+
- [GitHub Pages Documentation Generation](#github-pages-documentation-generation)
35+
- [Additional Features](#additional-features)
36+
- [Statusline](#statusline)
3637
- [Documentation](#documentation)
3738
- [Getting Started](#getting-started)
3839
- [Features](#features-1)
@@ -97,113 +98,6 @@ amplihack launch --with-proxy-config ./azure.env
9798
amplihack launch --checkout-repo owner/repo
9899
```
99100

100-
## Features
101-
102-
### Workflow Orchestration by Default
103-
104-
All prompts are automatically wrapped with `/amplihack:ultrathink` for maximum
105-
effectiveness. This enables:
106-
107-
- Multi-agent workflow orchestration
108-
- 13-step development workflow
109-
- Automated architecture, building, and testing
110-
- Philosophy compliance checking
111-
112-
**Benchmark results:** Amplihack without orchestration = vanilla Claude. The
113-
orchestration IS the value! See [benchmarking guide](docs/BENCHMARKING.md) for
114-
measuring performance.
115-
116-
**Opt-out for simple tasks:**
117-
118-
```sh
119-
# Skip orchestration with --no-ultrathink flag
120-
amplihack launch --no-ultrathink -- -p "simple prompt"
121-
122-
# Or use slash commands directly
123-
amplihack launch -- -p "/analyze src/file.py"
124-
```
125-
126-
**How it works:**
127-
128-
```sh
129-
# Before: Manual orchestration required
130-
amplihack launch -- -p "/amplihack:ultrathink implement feature"
131-
132-
# Now: Automatic orchestration (same result)
133-
amplihack launch -- -p "implement feature"
134-
```
135-
136-
### Goal-Seeking Agent Generator
137-
138-
**Create autonomous agents from simple prompts:**
139-
140-
```bash
141-
# Write your goal
142-
cat > my_goal.md <<'EOF'
143-
# Goal: Automated Code Review
144-
Review Python code and suggest improvements.
145-
EOF
146-
147-
# Generate agent
148-
amplihack new --file my_goal.md
149-
150-
# Run agent
151-
cd goal_agents/automated-code-review-agent
152-
python main.py
153-
```
154-
155-
**Features:**
156-
157-
- Generate agents in < 0.1 seconds
158-
- Automatic skill matching
159-
- Multi-phase execution planning
160-
- Standalone, distributable agents
161-
162-
**Learn more:** [Goal Agent Generator Guide](docs/GOAL_AGENT_GENERATOR_GUIDE.md)
163-
164-
### Profile Management
165-
166-
**Reduce token usage by 72% with profile-based component filtering:**
167-
168-
```bash
169-
# Set profile (outside Claude Code)
170-
export AMPLIHACK_PROFILE=amplihack://profiles/coding
171-
172-
# Install with filtering
173-
amplihack install
174-
# Result: Only 9/32 agents staged (72% reduction)
175-
176-
# Launch with filtering
177-
amplihack launch
178-
# Result: Focused environment for coding tasks
179-
180-
# To switch profiles: exit Claude, set new profile, restart
181-
```
182-
183-
**Built-in Profiles:**
184-
185-
- `all`: Full environment (32 agents, default)
186-
- `coding`: Development-focused (9 agents)
187-
- `research`: Investigation-focused (7 agents)
188-
189-
**Learn more:** [Profile Management Guide](docs/PROFILE_MANAGEMENT.md)
190-
191-
### GitHub Pages Documentation Generation
192-
193-
**Generate professional documentation sites automatically:**
194-
195-
- Auto-discovers content from `docs/`, `README.md`, and `.claude/commands/`
196-
- Three-pass validation ensures quality documentation
197-
- Safe gh-pages deployment with rollback support
198-
- Local preview server for testing
199-
- MkDocs + Material theme integration
200-
201-
**Learn more:**
202-
203-
- [Tutorial: Your First Documentation Site](docs/tutorials/first-docs-site.md)
204-
- [How-To: Generate GitHub Pages Sites](docs/howto/github-pages-generation.md)
205-
- [API Reference: GitHub Pages Module](docs/reference/github-pages-api.md)
206-
207101
## Core Concepts
208102

209103
### Workflow
@@ -363,7 +257,159 @@ customization instructions.
363257
| **xpia-defense** | Advanced threat detection |
364258
| **philosophy-guardian** | Philosophy compliance and simplicity validation |
365259

366-
## Statusline
260+
## Features
261+
262+
### Workflow Orchestration by Default
263+
264+
All prompts are automatically wrapped with `/amplihack:ultrathink` for maximum
265+
effectiveness. This enables:
266+
267+
- Multi-agent workflow orchestration
268+
- 13-step development workflow
269+
- Automated architecture, building, and testing
270+
- Philosophy compliance checking
271+
272+
**Benchmark results:** Amplihack without orchestration = vanilla Claude. The
273+
orchestration IS the value! See [benchmarking guide](docs/BENCHMARKING.md) for
274+
measuring performance.
275+
276+
**Opt-out for simple tasks:**
277+
278+
```sh
279+
# Skip orchestration with --no-ultrathink flag
280+
amplihack launch --no-ultrathink -- -p "simple prompt"
281+
282+
# Or use slash commands directly
283+
amplihack launch -- -p "/analyze src/file.py"
284+
```
285+
286+
**How it works:**
287+
288+
```sh
289+
# Before: Manual orchestration required
290+
amplihack launch -- -p "/amplihack:ultrathink implement feature"
291+
292+
# Now: Automatic orchestration (same result)
293+
amplihack launch -- -p "implement feature"
294+
```
295+
296+
### Goal-Seeking Agent Generator
297+
298+
**Create autonomous agents from simple prompts:**
299+
300+
```bash
301+
# Write your goal
302+
cat > my_goal.md <<'EOF'
303+
# Goal: Automated Code Review
304+
Review Python code and suggest improvements.
305+
EOF
306+
307+
# Generate agent
308+
amplihack new --file my_goal.md
309+
310+
# Run agent
311+
cd goal_agents/automated-code-review-agent
312+
python main.py
313+
```
314+
315+
**Features:**
316+
317+
- Generate agents in < 0.1 seconds
318+
- Automatic skill matching
319+
- Multi-phase execution planning
320+
- Standalone, distributable agents
321+
322+
**Learn more:** [Goal Agent Generator Guide](docs/GOAL_AGENT_GENERATOR_GUIDE.md)
323+
324+
### Profile Management
325+
326+
**Reduce token usage by 72% with profile-based component filtering:**
327+
328+
```bash
329+
# Set profile (outside Claude Code)
330+
export AMPLIHACK_PROFILE=amplihack://profiles/coding
331+
332+
# Install with filtering
333+
amplihack install
334+
# Result: Only 9/32 agents staged (72% reduction)
335+
336+
# Launch with filtering
337+
amplihack launch
338+
# Result: Focused environment for coding tasks
339+
340+
# To switch profiles: exit Claude, set new profile, restart
341+
```
342+
343+
**Built-in Profiles:**
344+
345+
- `all`: Full environment (32 agents, default)
346+
- `coding`: Development-focused (9 agents)
347+
- `research`: Investigation-focused (7 agents)
348+
349+
**Learn more:** [Profile Management Guide](docs/PROFILE_MANAGEMENT.md)
350+
351+
### GitHub Pages Documentation Generation
352+
353+
**Generate professional documentation sites automatically:**
354+
355+
- Auto-discovers content from `docs/`, `README.md`, and `.claude/commands/`
356+
- Three-pass validation ensures quality documentation
357+
- Safe gh-pages deployment with rollback support
358+
- Local preview server for testing
359+
- MkDocs + Material theme integration
360+
361+
**Learn more:**
362+
363+
- [Tutorial: Your First Documentation Site](docs/tutorials/first-docs-site.md)
364+
- [How-To: Generate GitHub Pages Sites](docs/howto/github-pages-generation.md)
365+
- [API Reference: GitHub Pages Module](docs/reference/github-pages-api.md)
366+
367+
### Additional Features
368+
369+
- **[Power-Steering](docs/reference/STATUSLINE.md#power-steering)** - AI-powered
370+
session guidance with intelligent redirect detection (🚦 indicator)
371+
- **[Auto Mode](docs/AUTO_MODE.md)** - Autonomous agentic loops for multi-turn
372+
workflows (`/amplihack:auto`)
373+
- **[Lock Mode](docs/reference/STATUSLINE.md#lock-mode)** - Continuous work mode
374+
without stopping (`/amplihack:lock`, `/amplihack:unlock`) (🔒 indicator)
375+
- **[Document-Driven Development](docs/document_driven_development/README.md)** -
376+
Systematic methodology for large features with documentation-first approach
377+
- **[Fault-Tolerant Workflows](CLAUDE.md#fault-tolerance-patterns)** - N-version
378+
programming, multi-agent debate, and cascade fallback patterns
379+
- **[Security Analysis](CLAUDE.md#key-commands)** - XPIA cross-prompt injection
380+
defense (`/amplihack:xpia`)
381+
- **[Neo4j Memory System](docs/AGENT_MEMORY_QUICKSTART.md)** - Persistent memory
382+
and knowledge graphs across sessions
383+
- **[Investigation Workflow](CLAUDE.md#investigation-workflow)** - Deep
384+
knowledge excavation with historical context
385+
- **[Skills System](.claude/skills/README.md)** - 54+ skills including PDF,
386+
XLSX, DOCX, PPTX, analysts, and workflow patterns
387+
- **[Fix Workflow](CLAUDE.md#key-commands)** - Rapid resolution of common error
388+
patterns (`/amplihack:fix`)
389+
- **[Reflection & Improvement](CLAUDE.md#key-commands)** - Session analysis and
390+
learning capture (`/amplihack:reflect`, `/amplihack:improve`)
391+
- **[Socratic Questioning](CLAUDE.md#key-commands)** - Challenge claims and
392+
clarify requirements (`/amplihack:socratic`)
393+
- **[Expert Panel](CLAUDE.md#key-commands)** - Multi-expert review with voting
394+
(`/amplihack:expert-panel`)
395+
- **[Knowledge Builder](CLAUDE.md#key-commands)** - Build comprehensive
396+
knowledge base (`/amplihack:knowledge-builder`)
397+
- **[Transcripts Management](CLAUDE.md#key-commands)** - Conversation transcript
398+
tracking (`/amplihack:transcripts`)
399+
- **[Modular Build](CLAUDE.md#key-commands)** - Self-contained modules with
400+
clear contracts (`/amplihack:modular-build`)
401+
- **[Pre-commit Diagnostics](CLAUDE.md#development-workflow-agents)** - Fix
402+
formatting, linting, type checking before push
403+
- **[CI Diagnostics](CLAUDE.md#development-workflow-agents)** - Monitor CI,
404+
diagnose failures, iterate until mergeable
405+
- **[Worktree Management](.claude/agents/amplihack/specialized/worktree-manager.md)** -
406+
Git worktree automation for parallel development
407+
- **[Session Logs](CLAUDE.md#working-philosophy)** - Comprehensive logging and
408+
decision records
409+
- **[Customization System](CLAUDE.md#key-commands)** - Manage user preferences
410+
(`/amplihack:customize`)
411+
412+
### Statusline
367413

368414
Real-time session information displayed at the bottom of Claude Code showing:
369415

0 commit comments

Comments
 (0)