Add orchestrator mode (cc-orc) with workhorse agents (#14)

* Add orchestrator mode with delegation-first system prompt and 4 workhorse agents

Introduces cc-orc alias for a slim orchestrator that decomposes tasks and
delegates to agents instead of doing implementation work directly. Four new
workhorse agents (investigator, implementer, tester, documenter) carry the
detailed execution discipline, code standards, and testing standards that
previously lived only in the monolithic main prompt. All agents enforce a
mandatory question surfacing protocol — they stop and return questions to the
orchestrator rather than making assumptions. The existing 17 specialist agents
and main-system-prompt.md remain unchanged.

* Fix 10 CodeRabbit review issues and stale documentation

Resolve all actionable, outside-diff, duplicate, and nitpick issues from
CodeRabbit review on PR #14, plus one discovered consistency issue.

Major: implementer PostToolUse hook (per-Edit) → Stop hook (once at end)
with 120s timeout. Tester Stop hook timeout 30s → 120s. Added cc-orc to
cc-tools discovery. Updated CLAUDE.md directory tree with missing files.
Fixed stale verify-no-regression.py comment. Added MD040 language
specifiers. Clarified orchestrator plan mode delegation scope and added
specialist catalog catch-all.

---------

Co-authored-by: AnExiledDev <AnExiledDev@users.noreply.github.com>

Author: AnExiledDev

.devcontainer/CHANGELOG.md

@@ -20,6 +20,28 @@
 #### Features
 - **devcontainer-bridge (dbr)** — Ports opened inside the container are now automatically discovered and forwarded to the host, even outside VS Code. Requires `dbr host-daemon` running on the host. See [devcontainer-bridge](https://github.com/bradleybeddoes/devcontainer-bridge)
 
+#### Orchestrator Mode
+- **`cc-orc` alias** — new Claude Code entry point using `orchestrator-system-prompt.md` for delegation-first operation; orchestrator decomposes tasks, delegates to agents, surfaces questions, and synthesizes results without performing direct implementation work
+- **`orchestrator-system-prompt.md`** — slim system prompt (~250 lines) containing only delegation model, agent catalog, question surfacing protocol, planning gates, spec enforcement, and action safety; all code standards, testing standards, and implementation details live in agent prompts
+
+#### Workhorse Agents
+- **`investigator`** — consolidated read-only research agent (sonnet) merging the domains of researcher, explorer, dependency-analyst, git-archaeologist, debug-logs, and perf-profiler; handles codebase search, web research, git forensics, dependency auditing, log analysis, and performance profiling
+- **`implementer`** — consolidated read-write implementation agent (opus, worktree) merging generalist, refactorer, and migrator; handles all code modifications with embedded code standards, execution discipline, and Stop hook regression testing
+- **`tester`** — enhanced test agent (opus, worktree) with full testing standards, framework-specific guidance, and Stop hook verification; creates and verifies test suites
+- **`documenter`** — consolidated documentation and specification agent (opus) merging doc-writer and spec-writer; handles README, API docs, docstrings, and the full spec lifecycle (create, refine, build, review, update, check)
+- **Question Surfacing Protocol** — all 4 workhorse agents carry an identical protocol requiring them to STOP and return `## BLOCKED: Questions` sections when hitting ambiguities, ensuring no assumptions are made without user input
+
+### Fixed
+
+#### CodeRabbit Review Fixes
+- **`implementer.md`** — changed PostToolUse hook (fires every Edit) to Stop hook (fires once at task end) with 120s timeout; prevents redundant test runs during multi-file tasks
+- **`tester.md`** — increased Stop hook timeout from 30s to 120s to accommodate larger test suites
+- **`setup-aliases.sh`** — added `cc-orc` to `cc-tools` discovery loop so it appears in tool audit
+- **`CLAUDE.md`** — added missing `keybindings.json`, `orchestrator-system-prompt.md`, and `writing-system-prompt.md` to directory structure tree
+- **`agent-system/README.md`** — updated `verify-no-regression.py` comment to list both consumers (implementer, refactorer); hyphenated "question-surfacing protocol"
+- **`orchestrator-system-prompt.md`** — clarified plan mode allows investigator delegation for research; added catch-all entry in selection criteria pointing to the full specialist catalog
+- **MD040 compliance** — added `text` language specifiers to 7 fenced code blocks across `investigator.md`, `tester.md`, and `documenter.md`
+
 ### Changed
 
 #### Skill Engine: Auto-Suggestion
@@ -40,6 +62,13 @@
 - Moved `.claude` directory from `/workspaces/.claude` to `~/.claude` (home directory)
 - Added Docker named volume for persistence across rebuilds (per-instance isolation via `${devcontainerId}`)
 - `CLAUDE_CONFIG_DIR` now defaults to `~/.claude`
+- `file-manifest.json` — added deployment entry for `orchestrator-system-prompt.md`
+- `setup-aliases.sh` — added `cc-orc` alias alongside existing `cc`, `claude`, `ccw`, `ccraw`
+- `CLAUDE.md` — documented `cc-orc` command and orchestrator system prompt in key configuration table
+
+#### Agent System
+- Agent count increased from 17 to 21 (4 workhorse + 17 specialist)
+- Agent-system README updated with workhorse agent table, per-agent hooks for implementer and tester, and updated plugin structure
 
 #### Authentication
 - Added `CLAUDE_AUTH_TOKEN` support in `.secrets` for long-lived tokens from `claude setup-token`

.devcontainer/CLAUDE.md

@@ -12,7 +12,10 @@ CodeForge devcontainer for AI-assisted development with Claude Code.
 │   ├── file-manifest.json     # Declarative config file deployment
 │   └── defaults/              # Source files deployed on start via file-manifest
 │       ├── settings.json      # Model, permissions, plugins, env vars
+│       ├── keybindings.json   # Keyboard shortcuts
 │       ├── main-system-prompt.md
+│       ├── orchestrator-system-prompt.md
+│       ├── writing-system-prompt.md
 │       ├── ccstatusline-settings.json  # Status bar widget layout
 │       └── rules/             # Deployed to .claude/rules/
 ├── features/                  # Custom devcontainer features
@@ -26,6 +29,7 @@ CodeForge devcontainer for AI-assisted development with Claude Code.
 |------|---------|
 | `config/defaults/settings.json` | Model, tokens, permissions, plugins, env vars |
 | `config/defaults/main-system-prompt.md` | System prompt defining assistant behavior |
+| `config/defaults/orchestrator-system-prompt.md` | Orchestrator mode prompt (delegation-first) |
 | `config/defaults/ccstatusline-settings.json` | Status bar widget layout (deployed to ~/.config/ccstatusline/) |
 | `config/file-manifest.json` | Controls which config files deploy and when |
 | `devcontainer.json` | Container definition: image, features, mounts |
@@ -68,6 +72,7 @@ git worktree add /workspaces/projects/.worktrees/<branch-name> -b <branch>
 | `cc` / `claude` | Run Claude Code with auto-configuration |
 | `ccraw` | Vanilla Claude Code (bypasses config) |
 | `ccw` | Claude Code with writing system prompt |
+| `cc-orc` | Claude Code in orchestrator mode (delegation-first) |
 | `ccms` | Search session history (project-scoped) |
 | `ccusage` / `ccburn` | Token usage analysis / burn rate |
 | `agent-browser` | Headless Chromium (Playwright-based) |
@@ -80,7 +85,7 @@ git worktree add /workspaces/projects/.worktrees/<branch-name> -b <branch>
 
 Declared in `settings.json` under `enabledPlugins`, auto-activated on start:
 
-- **agent-system** — 17 custom agents + built-in agent redirection
+- **agent-system** — 21 custom agents (4 workhorse + 17 specialist) + built-in agent redirection
 - **skill-engine** — 22 general coding skills + auto-suggestion
 - **spec-workflow** — 8 spec lifecycle skills + spec-reminder hook
 - **session-context** — Git state injection, TODO harvesting, commit reminders