claude-raid 0.2.4 → 0.2.5

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/README.md +173 -193
  2. package/package.json +1 -1
package/README.md CHANGED
@@ -1,89 +1,97 @@
1
- # claude-raid
2
-
3
- ```ansi
4
- [33m ═══════════════════════════════════════════════════════[0m
5
-
6
- [1;33m ██████╗██╗ █████╗ ██╗ ██╗██████╗ ███████╗[0m
7
- [1;33m ██╔════╝██║ ██╔══██╗██║ ██║██╔══██╗██╔════╝[0m
8
- [33m ██║ ██║ ███████║██║ ██║██║ ██║█████╗ [0m
9
- [33m ██║ ██║ ██╔══██║██║ ██║██║ ██║██╔══╝ [0m
10
- [33m ╚██████╗███████╗██║ ██║╚██████╔╝██████╔╝███████╗[0m
11
- [1;31m ╚═════╝╚══════╝╚═╝ ╚═╝ ╚═════╝ ╚═════╝ ╚══════╝[0m
12
- [1;31m ██████╗ █████╗ ██╗██████╗ [0m
13
- [1;31m ██╔══██╗██╔══██╗██║██╔══██╗[0m
14
- [31m ██████╔╝███████║██║██║ ██║[0m
15
- [31m ██╔══██╗██╔══██║██║██║ ██║[0m
16
- [31m ██║ ██║██║ ██║██║██████╔╝[0m
17
- [2;31m ╚═╝ ╚═╝╚═╝ ╚═╝╚═╝╚═════╝ [0m
18
-
19
- [90m Adversarial multi-agent development for Claude Code[0m
20
-
21
- [33m ⚔ ═══════════════════════════════════════════════════════ ⚔[0m
1
+ <div align="center">
2
+
3
+ ```
4
+ ══════════════════════════════════════════════════════════
5
+
6
+ ██████╗██╗ █████╗ ██╗ ██╗██████╗ ███████╗
7
+ ██╔════╝██║ ██╔══██╗██║ ██║██╔══██╗██╔════╝
8
+ ██║ ██║ ███████║██║ ██║██║ ██║█████╗
9
+ ██║ ██║ ██╔══██║██║ ██║██║ ██║██╔══╝
10
+ ╚██████╗███████╗██║ ██║╚██████╔╝██████╔╝███████╗
11
+ ╚═════╝╚══════╝╚═╝ ╚═╝ ╚═════╝ ╚═════╝ ╚══════╝
12
+ ██████╗ █████╗ ██╗██████╗
13
+ ██╔══██╗██╔══██╗██║██╔══██╗
14
+ ██████╔╝███████║██║██║ ██║
15
+ ██╔══██╗██╔══██║██║██║ ██║
16
+ ██║ ██║██║ ██║██║██████╔╝
17
+ ╚═╝ ╚═╝╚═╝ ╚═╝╚═╝╚═════╝
18
+
19
+ Round-based Adversarial Intelligence Dungeon
20
+ for Claude Code
21
+
22
+ ⚔ ══════════════════════════════════════════════════════════ ⚔
22
23
  ```
23
24
 
24
- Four AI agents work through a strict 4-phase workflow where every design decision, implementation choice, and code review is stress-tested by competing perspectives before it ships.
25
+ **Four AI agents. One shared dungeon. Every decision stress-tested before it ships.**
25
26
 
26
- Built for [Claude Code](https://claude.ai/claude-code). Zero dependencies. One command to install.
27
+ [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
28
+ [![Zero Dependencies](https://img.shields.io/badge/dependencies-0-brightgreen.svg)](#)
29
+ [![Node.js 18+](https://img.shields.io/badge/node-18%2B-blue.svg)](#prerequisites)
27
30
 
28
- Adapted from [obra/superpowers](https://github.com/obra/superpowers) by Jesse Vincent.
31
+ [Quick Start](#quick-start) &bull; [How It Works](#how-it-works) &bull; [The Team](#the-team) &bull; [Configuration](#configuration) &bull; [CLI Reference](#cli-reference)
32
+
33
+ </div>
29
34
 
30
35
  ---
31
36
 
32
37
  ## Quick Start
33
38
 
34
39
  ```bash
35
- # Install into any project
36
40
  npx claude-raid summon
41
+ ```
42
+
43
+ That's it. One command installs the full system into any project.
37
44
 
45
+ ```bash
38
46
  # Preview what gets installed (no changes made)
39
47
  npx claude-raid summon --dry-run
40
48
 
41
- # Start a tmux session, then start the Raid
49
+ # Start a raid
42
50
  tmux new-session -s raid
43
51
  claude --agent wizard
44
52
  ```
45
53
 
46
- Each agent gets its own tmux pane. You can click into any pane to talk directly to that agent. Describe your task and the Wizard takes over.
54
+ Each agent gets its own tmux pane. Click into any pane to observe or talk to that agent directly. Describe your task and the Wizard takes over.
47
55
 
48
56
  ### Prerequisites
49
57
 
50
58
  | Requirement | Why | Auto-configured? |
51
- |---|---|---|
59
+ |:--|:--|:--|
52
60
  | **Node.js** 18+ | Runs the installer | No |
53
61
  | **Claude Code** v2.1.32+ | Agent teams support | No |
54
62
  | **tmux** | Multi-pane agent display | No — `brew install tmux` |
55
63
  | **jq** | Hook config parsing | Pre-installed on macOS |
56
64
  | **teammateMode** | Set to `tmux` in `~/.claude.json` | Yes (wizard prompts) |
57
65
 
58
- **tmux is required for the multi-pane experience.** Each agent runs in its own pane so you can observe and interact with them independently. Without tmux, agents run in-process (single pane, cycle with Shift+Down).
66
+ > **tmux is required for the multi-pane experience.** Each agent runs in its own pane so you can observe and interact with them independently. Without tmux, agents run in-process (single pane, cycle with `Shift+Down`).
59
67
 
60
- The setup wizard checks all of these during `summon` and offers to fix what it can.
68
+ The setup wizard checks all prerequisites during `summon` and offers to fix what it can.
61
69
 
62
70
  ---
63
71
 
64
72
  ## How It Works
65
73
 
66
- You describe a task. The Wizard assesses complexity, recommends a mode, and opens the Dungeon — a shared knowledge artifact where agents pin verified findings throughout each phase.
74
+ You describe a task. The Wizard assesses complexity, recommends a mode, and opens the **Dungeon** — a shared artifact where agents pin verified findings throughout each phase.
67
75
 
68
76
  ```
69
- Phase 1: DESIGN Agents explore the problem from competing angles.
70
- Challenge each other's findings. Pin what survives.
71
- Wizard closes when the design is battle-tested.
77
+ DESIGN Agents explore the problem from competing angles.
78
+ Challenge each other's findings. Pin what survives.
79
+ Wizard closes when the design is battle-tested.
72
80
 
73
- Phase 2: PLAN Decompose the design into testable tasks.
74
- Fight over naming, ordering, coverage, compliance.
75
- Pin the agreed task list.
81
+ PLAN Decompose the design into testable tasks.
82
+ Fight over naming, ordering, coverage, compliance.
83
+ Pin the agreed task list.
76
84
 
77
- Phase 3: IMPLEMENTATION One agent builds each task (TDD enforced).
78
- The others attack the implementation directly.
79
- Every task earns approval before moving on.
85
+ IMPLEMENTATION One agent builds each task (TDD enforced).
86
+ The others attack the implementation directly.
87
+ Every task earns approval before moving on.
80
88
 
81
- Phase 4: REVIEW Independent reviews against design and plan.
82
- Agents fight over findings AND missing findings.
83
- Critical and Important issues must be fixed.
89
+ REVIEW Independent reviews against design and plan.
90
+ Agents fight over findings AND missing findings.
91
+ Critical and Important issues must be fixed.
84
92
 
85
- FINISHING Agents debate completeness. Wizard presents
86
- merge options: merge, PR, keep branch, or discard.
93
+ FINISHING Agents debate completeness. Wizard presents
94
+ merge options: merge, PR, keep branch, or discard.
87
95
  ```
88
96
 
89
97
  No phase is skipped. No work passes unchallenged.
@@ -92,30 +100,29 @@ No phase is skipped. No work passes unchallenged.
92
100
 
93
101
  ## The Team
94
102
 
95
- ### Wizard — Dungeon Master
96
-
97
- Thinks 5 times before speaking. Opens phases, dispatches the team with precise angles, observes silently (90% of the time), and closes each phase with a binding ruling citing Dungeon evidence. The Wizard doesn't write code — it ensures the process produces the best possible outcome.
98
-
99
- ### Warrior Stress Tester
100
-
101
- Rips things apart. Race conditions, null input, scale, memory pressure — nothing passes unchecked. Demands evidence, proposes counter-examples, and pushes until things break. Concedes instantly when proven wrong.
102
-
103
- ### Archer Pattern Seeker
104
-
105
- Finds what brute force misses. Naming mismatches, violated conventions, design drift. Traces ripple effects changing X in module A silently breaks Y in module C through an implicit contract in Z. Every finding includes the exact location and the exact consequence.
106
-
107
- ### Rogue — Assumption Destroyer
108
-
109
- Thinks like a malicious user, a failing network, a corrupted database, a race condition at 3 AM. Constructs the exact sequence of events that turns a minor oversight into a critical failure. Every finding is a concrete attack scenario, not a theoretical concern.
103
+ <table>
104
+ <tr>
105
+ <td width="25%" align="center"><h3>Wizard</h3><em>Dungeon Master</em></td>
106
+ <td width="25%" align="center"><h3>Warrior</h3><em>Stress Tester</em></td>
107
+ <td width="25%" align="center"><h3>Archer</h3><em>Pattern Seeker</em></td>
108
+ <td width="25%" align="center"><h3>Rogue</h3><em>Assumption Destroyer</em></td>
109
+ </tr>
110
+ <tr>
111
+ <td valign="top">Thinks 5 times before speaking. Opens phases, dispatches the team, observes silently, and closes with binding rulings citing Dungeon evidence. <strong>Never writes code.</strong></td>
112
+ <td valign="top">Race conditions, null input, scale, memory pressure — nothing passes unchecked. Demands evidence, proposes counter-examples, pushes until things break. Concedes instantly when proven wrong.</td>
113
+ <td valign="top">Finds what brute force misses. Naming mismatches, violated conventions, design drift. Traces ripple effects across modules through implicit contracts.</td>
114
+ <td valign="top">Thinks like a malicious user, a failing network, a corrupted database. Constructs the exact sequence of events that turns a minor oversight into a critical failure.</td>
115
+ </tr>
116
+ </table>
110
117
 
111
118
  ---
112
119
 
113
120
  ## Modes
114
121
 
115
- Not every task needs all four agents. The Wizard recommends a mode based on complexity, or you can override directly.
122
+ Not every task needs all four agents. The Wizard recommends a mode based on complexity or you can override directly.
116
123
 
117
124
  | | Full Raid | Skirmish | Scout |
118
- |---|---|---|---|
125
+ |:--|:--|:--|:--|
119
126
  | **Agents** | 3 + Wizard | 2 + Wizard | 1 + Wizard |
120
127
  | **Design** | Full adversarial | Lightweight | Inline |
121
128
  | **Planning** | Full adversarial | Merged with design | Inline |
@@ -131,30 +138,32 @@ Not every task needs all four agents. The Wizard recommends a mode based on comp
131
138
 
132
139
  Override the Wizard's recommendation: *"Full Raid this"*, *"Skirmish this bugfix"*, *"Scout this"*.
133
140
 
134
- The Wizard can escalate mid-task (Scout to Skirmish, Skirmish to Full Raid) with your approval. It cannot de-escalate without asking.
141
+ > The Wizard can escalate mid-task (Scout to Skirmish, Skirmish to Full Raid) with your approval. It cannot de-escalate without asking.
135
142
 
136
143
  ---
137
144
 
138
145
  ## The Dungeon
139
146
 
140
- The Dungeon (`.claude/raid-dungeon.md`) is the team's shared knowledge artifact — a curated board where agents pin verified findings during each phase.
147
+ The Dungeon is the team's shared knowledge artifact — a curated board where agents pin verified findings during each phase.
141
148
 
142
149
  **What goes in:** Findings that survived challenge from 2+ agents, active unresolved battles, key decisions, escalation points.
143
150
 
144
- **What stays in conversation:** The back-and-forth of challenges and roasts, exploratory thinking, concessions. The conversation is the sparring ring. The Dungeon is the scoreboard.
145
-
146
- **Lifecycle:** Wizard creates it when opening a phase, agents pin findings with `DUNGEON:` prefix, Wizard archives it when closing (`raid-dungeon-phase-N.md`), and all Dungeon files are cleaned up when the session ends.
151
+ **What stays in conversation:** The back-and-forth of challenges, exploratory thinking, concessions. The conversation is the sparring ring. The Dungeon is the scoreboard.
147
152
 
148
- Agents interact directly `@Name` mentions, building on each other's discoveries, roasting weak analysis. The Wizard observes silently, intervening only on destructive loops, drift, deadlock, or misinformation.
153
+ **Lifecycle:** The Wizard creates it when opening a phase. Agents pin findings with `DUNGEON:` prefix. The Wizard archives it when closing. All Dungeon files are cleaned up when the session ends.
149
154
 
150
155
  ---
151
156
 
152
157
  ## What Gets Installed
153
158
 
159
+ <details>
160
+ <summary><strong>Full file tree</strong></summary>
161
+
154
162
  ```
155
163
  .claude/
156
164
  ├── raid.json # Project config (auto-detected, editable)
157
- ├── raid-rules.md # 17 team rules across 3 pillars (editable)
165
+ ├── party-rules.md # Party agent rules (editable)
166
+ ├── dungeon-master-rules.md # Wizard rules (editable)
158
167
  ├── settings.json # Hooks merged with existing (backup created)
159
168
  ├── agents/
160
169
  │ ├── wizard.md # Dungeon master
@@ -165,11 +174,8 @@ Agents interact directly — `@Name` mentions, building on each other's discover
165
174
  │ ├── raid-lib.sh # Shared config and session state
166
175
  │ ├── raid-session-start.sh # Session activation
167
176
  │ ├── raid-session-end.sh # Archive and cleanup
168
- │ ├── raid-stop.sh # Phase transition backup
169
177
  │ ├── raid-pre-compact.sh # Pre-compaction backup
170
178
  │ ├── raid-task-created.sh # Task subject validation
171
- │ ├── raid-task-completed.sh # Test evidence gate
172
- │ ├── raid-teammate-idle.sh # Idle agent nudge
173
179
  │ ├── validate-commit.sh # Conventional commits + test gate
174
180
  │ ├── validate-write-gate.sh # Design doc before implementation
175
181
  │ ├── validate-file-naming.sh # Naming convention enforcement
@@ -178,112 +184,89 @@ Agents interact directly — `@Name` mentions, building on each other's discover
178
184
  │ ├── validate-browser-tests-exist.sh # Playwright test detection
179
185
  │ └── validate-browser-cleanup.sh # Browser process cleanup
180
186
  └── skills/
181
- ├── raid-protocol/ # Session lifecycle and team rules
182
- ├── raid-design/ # Phase 1: adversarial exploration
183
- ├── raid-implementation-plan/ # Phase 2: task decomposition
184
- ├── raid-implementation/ # Phase 3: TDD with direct challenge
185
- ├── raid-review/ # Phase 4: independent review + fighting
186
- ├── raid-finishing/ # Completeness debate + merge options
187
+ ├── raid-init/ # Quest selection and session setup
188
+ ├── raid-canonical-protocol/ # Canonical Quest rules and signals
189
+ ├── raid-canonical-prd/ # Phase 1: PRD creation
190
+ ├── raid-canonical-design/ # Phase 2: Adversarial design
191
+ ├── raid-canonical-implementation-plan/ # Phase 3: Task decomposition
192
+ ├── raid-canonical-implementation/ # Phase 4: TDD + cross-testing
193
+ ├── raid-canonical-review/ # Phase 5: Pinning + fixing
194
+ ├── raid-wrap-up/ # Phase 6: Storyboard + PR
187
195
  ├── raid-tdd/ # RED-GREEN-REFACTOR enforcement
188
- ├── raid-debugging/ # Root-cause investigation
189
196
  ├── raid-verification/ # Evidence-before-claims gate
190
- ├── raid-git-worktrees/ # Isolated workspace creation
191
- ├── raid-browser/ # Browser startup discovery
197
+ ├── raid-debugging/ # Root-cause investigation
198
+ ├── raid-browser/ # Browser orchestration
192
199
  ├── raid-browser-playwright/ # Playwright test authoring
193
- └── raid-browser-chrome/ # Live browser inspection
200
+ └── raid-browser-chrome/ # Live Chrome inspection
194
201
  ```
195
202
 
196
- Dungeon files (`raid-dungeon.md`, `raid-dungeon-phase-*.md`) and session state (`raid-session`) are created at runtime during Raid sessions and cleaned up automatically.
203
+ </details>
204
+
205
+ Runtime files (dungeon artifacts, session state) are created during Raid sessions and cleaned up automatically.
197
206
 
198
207
  ---
199
208
 
200
209
  ## Hooks
201
210
 
202
- Hooks enforce workflow discipline automatically. They split into two categories:
211
+ Hooks enforce workflow discipline automatically. They only activate during Raid sessions — they never interfere with normal coding.
203
212
 
204
- ### Lifecycle Hooks
205
-
206
- Manage session state without manual intervention. These activate and deactivate as sessions begin and end.
213
+ ### Lifecycle
207
214
 
208
215
  | Hook | Trigger | Purpose |
209
- |---|---|---|
210
- | **raid-session-start** | SessionStart | Activates Raid workflow, checks Vault for past quests |
211
- | **raid-session-end** | SessionEnd | Archives Dungeon, drafts Vault entry, removes session |
212
- | **raid-stop** | Stop | Backs up Dungeon on phase transitions |
213
- | **raid-pre-compact** | PreCompact | Backs up Dungeon before message compaction |
214
- | **raid-task-created** | TaskCreated | Validates task subjects are meaningful |
215
- | **raid-task-completed** | TaskCompleted | Blocks task completion without test evidence |
216
- | **raid-teammate-idle** | TeammateIdle | Nudges idle agents to participate |
216
+ |:--|:--|:--|
217
+ | `raid-session-start` | SessionStart | Activates Raid workflow, checks Vault for past quests |
218
+ | `raid-session-end` | SessionEnd | Archives Dungeon, drafts Vault entry, removes session |
219
+ | `raid-pre-compact` | PreCompact | Backs up Dungeon before message compaction |
220
+ | `raid-task-created` | TaskCreated | Validates task subjects are meaningful |
217
221
 
218
222
  ### Quality Gates
219
223
 
220
- Enforce code standards and workflow compliance. Quality gates only activate during Raid sessions — they won't interfere with normal coding.
221
-
222
224
  | Hook | Trigger | Purpose |
223
- |---|---|---|
224
- | **validate-commit** | PreToolUse (Bash) | Conventional commit format + test gate before commits |
225
- | **validate-write-gate** | PreToolUse (Write) | Blocks implementation files before design doc exists |
226
- | **validate-file-naming** | PostToolUse (Write) | Enforces naming convention (kebab-case, snake_case, etc.) |
227
- | **validate-no-placeholders** | PostToolUse (Write) | Blocks TBD/TODO/FIXME in specs and plans |
228
- | **validate-dungeon** | PostToolUse (Write) | Requires 2+ agents verified on Dungeon pins |
229
- | **validate-browser-tests-exist** | PreToolUse (Bash) | Checks Playwright tests exist before commits |
230
- | **validate-browser-cleanup** | PostToolUse (Bash) | Verifies browser processes cleaned up properly |
231
-
232
- All hooks are POSIX-compatible, read configuration from `raid.json`, and use `#claude-raid` markers to avoid collisions with your existing hooks.
225
+ |:--|:--|:--|
226
+ | `validate-commit` | PreToolUse (Bash) | Conventional commit format + test gate |
227
+ | `validate-write-gate` | PreToolUse (Write) | Blocks implementation before design doc exists |
228
+ | `validate-file-naming` | PostToolUse (Write) | Enforces naming conventions |
229
+ | `validate-no-placeholders` | PostToolUse (Write) | Blocks TBD/TODO/FIXME in specs and plans |
230
+ | `validate-dungeon` | PostToolUse (Write) | Requires 2+ agents verified on Dungeon pins |
231
+ | `validate-browser-tests-exist` | PreToolUse (Bash) | Checks Playwright tests exist before commits |
232
+ | `validate-browser-cleanup` | PostToolUse (Bash) | Verifies browser processes cleaned up |
233
233
 
234
- **Exit codes:** `0` = pass, `2` = block with message (agent sees the error and must fix it).
234
+ All hooks are POSIX-compatible, read config from `raid.json`, and use `#claude-raid` markers to avoid collisions with your existing hooks. Exit code `0` = pass, `2` = block with message.
235
235
 
236
236
  ---
237
237
 
238
238
  ## Skills
239
239
 
240
- 13 specialized skills guide agent behavior at each stage of the workflow.
240
+ 14 specialized skills guide agent behavior across the workflow.
241
241
 
242
242
  ### Phase Skills
243
243
 
244
244
  | Skill | Purpose |
245
- |---|---|
246
- | **raid-protocol** | Master orchestration. Session lifecycle, team composition, modes, rules, reference tables. Loaded at session start. |
247
- | **raid-design** | Phase 1. Agents independently explore from assigned angles, challenge findings, pin discoveries. Produces a battle-tested design spec. |
248
- | **raid-implementation-plan** | Phase 2. Decompose design into testable tasks. Agents debate naming, ordering, coverage, compliance. |
249
- | **raid-implementation** | Phase 3. One implements (TDD), others attack. Rotate implementer per task. No task passes without all issues resolved. |
250
- | **raid-review** | Phase 4. Independent reviews against design and plan. Issues classified as Critical (must fix), Important (must fix), or Minor (note). |
251
- | **raid-finishing** | Completeness debate. Agents argue whether work is done. Four merge options: merge, PR, keep branch, discard. |
245
+ |:--|:--|
246
+ | `raid-init` | Quest selection, greeting, session setup |
247
+ | `raid-canonical-protocol` | Canonical Quest rules, signals, phase gates |
248
+ | `raid-canonical-prd` | Phase 1 PRD creation (optional) |
249
+ | `raid-canonical-design` | Phase 2 Adversarial design exploration |
250
+ | `raid-canonical-implementation-plan` | Phase 3 Task decomposition |
251
+ | `raid-canonical-implementation` | Phase 4 TDD implementation + cross-testing |
252
+ | `raid-canonical-review` | Phase 5 — Pinning, fixing, black cards |
253
+ | `raid-wrap-up` | Phase 6 — Storyboard, PR, vault archival |
252
254
 
253
255
  ### Discipline Skills
254
256
 
255
257
  | Skill | Purpose |
256
- |---|---|
257
- | **raid-tdd** | Strict RED-GREEN-REFACTOR. No production code before a failing test. Challengers attack test quality. |
258
- | **raid-debugging** | Competing hypotheses in parallel. No fixes without confirmed root cause. 3+ failed fixes triggers architecture rethink. |
259
- | **raid-verification** | Evidence before assertions. Fresh test run required. Forbidden phrases without evidence: "done", "working", "fixed". |
260
- | **raid-git-worktrees** | Isolated workspace creation with safety verification and clean test baseline. |
258
+ |:--|:--|
259
+ | `raid-tdd` | Strict RED-GREEN-REFACTOR. No production code before a failing test. |
260
+ | `raid-debugging` | Competing hypotheses in parallel. No fixes without confirmed root cause. |
261
+ | `raid-verification` | Evidence before assertions. Fresh test run required before any completion claim. |
261
262
 
262
263
  ### Browser Skills
263
264
 
264
265
  | Skill | Purpose |
265
- |---|---|
266
- | **raid-browser** | Browser startup discovery. Detects dev server, auth requirements, startup steps. |
267
- | **raid-browser-playwright** | Playwright MCP test authoring with network and console assertions. |
268
- | **raid-browser-chrome** | Live browser inspection via Claude-in-Chrome MCP. Angle-driven investigation. |
269
-
270
- ---
271
-
272
- ## Team Rules
273
-
274
- 17 non-negotiable rules organized into 3 pillars. Stored in `.claude/raid-rules.md` (editable).
275
-
276
- ### Pillar 1: Intellectual Honesty
277
-
278
- Every claim has evidence you gathered yourself. If you haven't read the code or run the command this turn, you don't know what it says. Never fabricate evidence, certainty, or findings.
279
-
280
- ### Pillar 2: Zero Ego Collaboration
281
-
282
- When proven wrong, concede instantly. Defend with evidence, never with authority. A teammate catching your mistake is a gift. Build on each other's work — the best findings come from combining perspectives.
283
-
284
- ### Pillar 3: Discipline and Efficiency
285
-
286
- Maximum effort on every task. Every interaction carries work forward. Agents talk directly to each other — the Wizard is not a relay. Escalate to the Wizard only after you've tried to resolve it yourself.
266
+ |:--|:--|
267
+ | `raid-browser` | Browser startup discovery dev server, auth, startup steps |
268
+ | `raid-browser-playwright` | Playwright MCP test authoring with network and console assertions |
269
+ | `raid-browser-chrome` | Live browser inspection via Claude-in-Chrome MCP |
287
270
 
288
271
  ---
289
272
 
@@ -323,7 +306,7 @@ Maximum effort on every task. Every interaction carries work forward. Agents tal
323
306
  ### Auto-Detection
324
307
 
325
308
  | Marker File | Language | Test Command | Lint Command |
326
- |---|---|---|---|
309
+ |:--|:--|:--|:--|
327
310
  | `package.json` | JavaScript | `npm test` | `npm run lint` |
328
311
  | `Cargo.toml` | Rust | `cargo test` | `cargo clippy` |
329
312
  | `pyproject.toml` | Python | `pytest` / `poetry run pytest` | `ruff check .` |
@@ -332,10 +315,11 @@ Maximum effort on every task. Every interaction carries work forward. Agents tal
332
315
 
333
316
  Commands are extracted from your project files (e.g., `scripts.test` in `package.json`). Package manager is auto-detected (npm, pnpm, yarn, bun, uv, poetry). Edit `raid.json` to override any value.
334
317
 
335
- ### Configuration Reference
318
+ <details>
319
+ <summary><strong>Full configuration reference</strong></summary>
336
320
 
337
321
  | Key | Default | Description |
338
- |---|---|---|
322
+ |:--|:--|:--|
339
323
  | `project.testCommand` | auto-detected | Command to run tests |
340
324
  | `project.lintCommand` | auto-detected | Command to run linting |
341
325
  | `project.buildCommand` | auto-detected | Command to build |
@@ -349,6 +333,8 @@ Commands are extracted from your project files (e.g., `scripts.test` in `package
349
333
  | `raid.defaultMode` | `full` | Default mode: `full`, `skirmish`, `scout` |
350
334
  | `raid.lifecycle.testWindowMinutes` | `10` | Max age (minutes) of test run for verification |
351
335
 
336
+ </details>
337
+
352
338
  ### Browser Testing
353
339
 
354
340
  When a browser framework is detected (Next.js, Vite, Angular, etc.), a `browser` section is added to `raid.json`:
@@ -366,25 +352,25 @@ When a browser framework is detected (Next.js, Vite, Angular, etc.), a `browser`
366
352
  }
367
353
  ```
368
354
 
369
- This enables browser-specific hooks and skills — Playwright test detection, browser process cleanup, and live browser inspection during reviews.
355
+ This enables browser-specific hooks and skills — Playwright test detection, browser process cleanup, and live inspection during reviews.
370
356
 
371
357
  ---
372
358
 
373
- ## CLI Commands
359
+ ## CLI Reference
374
360
 
375
- | Command | Purpose |
376
- |---|---|
377
- | `claude-raid summon` | Install Raid into your project |
378
- | `claude-raid summon --dry-run` | Preview what would be installed |
379
- | `claude-raid update` | Upgrade hooks, skills, and rules to latest version |
380
- | `claude-raid dismantle` | Remove all Raid files, restore original settings |
381
- | `claude-raid heal` | Check environment health, show reference card |
361
+ | Command | Alias | Purpose |
362
+ |:--|:--|:--|
363
+ | `claude-raid summon` | `init` | Install Raid into your project |
364
+ | `claude-raid update` | | Upgrade hooks, skills, and rules to latest |
365
+ | `claude-raid dismantle` | `remove` | Remove all Raid files, restore original settings |
366
+ | `claude-raid heal` | `doctor` | Check environment health |
382
367
 
383
- Old names (`init`, `remove`, `doctor`) still work as aliases.
368
+ <details>
369
+ <summary><strong>Command details</strong></summary>
384
370
 
385
371
  ### `summon`
386
372
 
387
- Installs the full Raid system into your project. Auto-detects project type, copies agents/hooks/skills, generates `raid.json`, merges settings (with backup), and runs the setup wizard.
373
+ Installs the full Raid system. Auto-detects project type, copies agents/hooks/skills, generates `raid.json`, merges settings (with backup), and runs the setup wizard.
388
374
 
389
375
  - Never overwrites existing files — customized agents are preserved
390
376
  - Idempotent — safe to run multiple times
@@ -392,15 +378,17 @@ Installs the full Raid system into your project. Auto-detects project type, copi
392
378
 
393
379
  ### `update`
394
380
 
395
- Upgrades hooks, skills, and `raid-rules.md` to the latest version. Skips customized agents (warns which ones were preserved). Does not touch `raid.json` — your project config stays intact.
381
+ Upgrades hooks, skills, and rules to the latest version. Skips customized agents (warns which ones were preserved). Does not touch `raid.json`.
396
382
 
397
383
  ### `dismantle`
398
384
 
399
- Removes all Raid agents, hooks, skills, and config files. Restores `settings.json` from the backup created during install. Preserves non-Raid files in `.claude/`.
385
+ Removes all Raid agents, hooks, skills, and config files. Restores `settings.json` from the backup created during install.
400
386
 
401
387
  ### `heal`
402
388
 
403
- Checks Node.js, Claude Code, jq, teammateMode, and split-pane support. Offers to fix missing configuration interactively. Shows the "How It Works" reference card with modes, phases, hook enforcement, and controls.
389
+ Checks Node.js, Claude Code, jq, teammateMode, and split-pane support. Offers to fix missing configuration interactively.
390
+
391
+ </details>
404
392
 
405
393
  ---
406
394
 
@@ -409,63 +397,55 @@ Checks Node.js, Claude Code, jq, teammateMode, and split-pane support. Offers to
409
397
  **tmux pane navigation (recommended):**
410
398
 
411
399
  | Action | How |
412
- |---|---|
413
- | **Switch to agent pane** | Click the pane, or `Ctrl+B` then arrow key |
414
- | **Talk to an agent** | Click their pane and type |
415
- | **View all agents** | All panes visible simultaneously |
400
+ |:--|:--|
401
+ | Switch to agent pane | Click the pane, or `Ctrl+B` then arrow key |
402
+ | Talk to an agent | Click their pane and type |
403
+ | View all agents | All panes visible simultaneously |
416
404
 
417
405
  **In-process mode (no tmux):**
418
406
 
419
407
  | Shortcut | Action |
420
- |---|---|
421
- | **Shift+Down** | Cycle through teammates |
422
- | **Enter** | View a teammate's session |
423
- | **Escape** | Interrupt a teammate's turn |
424
- | **Ctrl+T** | Toggle the shared task list |
425
-
426
- ### Starting a Raid session
427
-
428
- ```bash
429
- # Always start tmux first for multi-pane
430
- tmux new-session -s raid
431
-
432
- # Then start the Wizard inside tmux
433
- claude --agent wizard
434
- ```
435
-
436
- The Wizard creates a team and spawns agents — each gets its own tmux pane automatically. If you're not inside tmux, agents fall back to in-process mode (single pane).
408
+ |:--|:--|
409
+ | `Shift+Down` | Cycle through teammates |
410
+ | `Enter` | View a teammate's session |
411
+ | `Escape` | Interrupt a teammate's turn |
412
+ | `Ctrl+T` | Toggle the shared task list |
437
413
 
438
414
  ---
439
415
 
440
- ## Non-Invasive Design
441
-
442
- The Raid is a tool in your toolkit, not your project's operating system.
416
+ ## Design Principles
443
417
 
444
- - **Never touches your `CLAUDE.md`** your project instructions stay yours
445
- - **Merges settings** alongside your existing config, with automatic backup
446
- - **Won't overwrite** existing agents, hooks, or skills that share a name
447
- - **Session-scoped hooks** — workflow hooks only activate during Raid sessions, never during normal coding
448
- - **Clean removal** restores your original `settings.json` from backup
449
- - **Zero npm dependencies** — pure Node.js stdlib, fast `npx` cold-start
418
+ - **Non-invasive** — Never touches your `CLAUDE.md`. Merges settings alongside your existing config with automatic backup. Clean removal restores originals.
419
+ - **Session-scoped** Quality gate hooks only activate during Raid sessions. They never interfere with normal coding.
420
+ - **Zero dependencies** Pure Node.js stdlib. Fast `npx` cold-start.
421
+ - **Safe by default** — Never overwrites existing files. Customized agents are always preserved.
450
422
 
451
423
  ---
452
424
 
453
425
  ## Inherited from Superpowers
454
426
 
455
- The Raid inherits and adapts the [Superpowers](https://github.com/obra/superpowers) behavioral harness:
427
+ Adapted from [obra/superpowers](https://github.com/obra/superpowers) by Jesse Vincent.
456
428
 
457
429
  | Concept | How the Raid Uses It |
458
- |---|---|
430
+ |:--|:--|
459
431
  | HARD-GATEs | No code before design approval. No implementation before plan approval. |
460
432
  | TDD Iron Law | No production code without a failing test first. Enforced in all modes. |
461
433
  | Verification Iron Law | No completion claims without fresh test evidence. |
462
434
  | No Placeholders | Specs and plans must contain complete content, not "TBD" or "implement later". |
463
435
  | Conventional Commits | Enforced via hook: `type(scope): description`. |
464
436
 
465
- **What's different:** Superpowers uses a single agent with subagent delegation. The Raid uses 4 agents in an agent team with adversarial cross-testing, direct interaction, and collaborative learning. Agents talk to each other (not through the Wizard), pin verified findings to a shared Dungeon, and self-organize within phases. Every decision is stress-tested from multiple angles before it passes.
437
+ **What's different:** Superpowers uses a single agent with subagent delegation. The Raid uses 4 persistent agents with adversarial cross-testing and direct interaction. Agents talk to each other, pin verified findings to a shared Dungeon, and self-organize within phases. Every decision is stress-tested from multiple angles before it passes.
466
438
 
467
439
  ---
468
440
 
469
441
  ## License
470
442
 
471
443
  MIT
444
+
445
+ ---
446
+
447
+ <div align="center">
448
+
449
+ Built for [Claude Code](https://claude.ai/claude-code).
450
+
451
+ </div>
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "claude-raid",
3
- "version": "0.2.4",
3
+ "version": "0.2.5",
4
4
  "type": "commonjs",
5
5
  "description": "Adversarial multi-agent development system for Claude Code",
6
6
  "author": "Pedro Picardi",