codeforge-dev 1.5.8 → 1.7.0

package/.devcontainer/.env

@@ -21,5 +21,6 @@ SETUP_PLUGINS=true
 SETUP_UPDATE_CLAUDE=true
 
 # Plugin blacklist (comma-separated plugin names to skip during auto-install)
-# Example: PLUGIN_BLACKLIST="workflow-enhancer,auto-formatter"
+# workflow-enhancer: Not fully implemented yet — scaffolding only
+# planning-reminder: Redundant with Claude Code v2.1+ which auto-switches to plan mode
 PLUGIN_BLACKLIST="workflow-enhancer,planning-reminder"

package/.devcontainer/CHANGELOG.md

@@ -1,5 +1,61 @@
 # CodeForge Devcontainer Changelog
 
+## [v1.7.0] - 2026-02-08
+
+### Added
+
+- **ccburn feature** — new devcontainer feature for visual token burn rate tracking with shell aliases and statusline wrapper
+- **Session resume widget** — ccstatusline displays copyable `cc --resume {sessionId}` command on line 5
+- **Burn rate widget** — ccstatusline line 6 shows live ccburn compact output with pace indicators (session/weekly/sonnet limits)
+- **17 custom agent definitions** — code-directive plugin now includes specialized agents: architect, bash-exec, claude-guide, debug-logs, dependency-analyst, doc-writer, explorer, generalist, git-archaeologist, migrator, perf-profiler, refactorer, researcher, security-auditor, spec-writer, statusline-config, test-writer
+- **6 new skills** — claude-agent-sdk, git-forensics, performance-profiling, refactoring-patterns, security-checklist, specification-writing
+- **Agent redirect hook** — `redirect-builtin-agents.py` (PreToolUse/Task) transparently swaps built-in agent types (Explore→explorer, Plan→architect, etc.) to enhanced custom agents
+- **Readonly bash guard** — `guard-readonly-bash.py` blocks write operations for read-only agents
+- **Regression test hooks** — `verify-no-regression.py` (PostToolUse for refactorer) and `verify-tests-pass.py` (Stop for test-writer)
+- **REVIEW-RUBRIC.md** — quality standards document for agent/skill development
+- **Keybindings configuration** — new `config/keybindings.json` with schema support
+- **VS Code terminal passthrough** — `Ctrl+P` and `Ctrl+F` pass through to Claude Code via `terminal.integrated.commandsToSkipShell`
+- **claude-agent-sdk skill** — new code-directive skill for Claude Agent SDK TypeScript integration
+- **OVERWRITE_CONFIG documentation** — documented ephemeral settings behavior
+- **Project Manager integration** — `setup-projects.sh` auto-detects projects under `/workspaces/`, watches for changes via inotifywait, maintains `projects.json`
+- **Claude config symlink** — `setup-symlink-claude.sh` symlinks `~/.claude` → `$CLAUDE_CONFIG_DIR` for third-party tool compatibility
+- **Project Manager VS Code extension** — `alefragnani.project-manager` added to devcontainer
+
+### Changed
+
+- **ccstatusline layout** — expanded from 3→6 lines (13→16 widgets), reorganized into logical groups (core metrics, tokens, git, session, totals, burn rate)
+- **ccstatusline version** — bumped from 1.0.0 to 1.1.0
+- **Plugin declarations centralized** — all 9 marketplace plugins declared in `enabledPlugins` in `config/settings.json`
+- **setup-plugins.sh cache sync** — re-added plugin install loop to sync cache from source on every container start; added `.env` fallback so `PLUGIN_BLACKLIST` works on standalone invocation
+- **Feature-level config synced** — `features/claude-code/config/settings.json` mirrors main config (model → `claude-opus-4-6`, `MAX_THINKING_TOKENS` → `63999`, `cleanupPeriodDays` → `60`, all env vars)
+- **8 new env vars** — `CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY`, `CLAUDE_CODE_MAX_RETRIES`, `BASH_MAX_OUTPUT_LENGTH`, `TASK_MAX_OUTPUT_LENGTH`, `CLAUDE_CODE_PLAN_MODE_INTERVIEW_PHASE`, `CLAUDE_CODE_PLAN_V2_AGENT_COUNT`, `CLAUDE_CODE_PLAN_MODE_REQUIRED`, `CLAUDE_CODE_FORCE_GLOBAL_CACHE`
+- **setup-config.sh** — added `chown` for correct ownership; added keybindings.json to copy pipeline
+- **setup-aliases.sh** — added idempotency guard
+- **TMPDIR consistency** — `setup-update-claude.sh` and `ccstatusline/install.sh` use `${TMPDIR:-/tmp}`
+- **installsAfter references** — mcp-qdrant and mcp-reasoner updated from `./features/claude-code` to `ghcr.io/anthropics/devcontainer-features/claude-code:1`
+- **code-directive hooks.json** — added PreToolUse/Task hook for agent redirection
+- **Auto-linter timeout** — pyright reduced from 55s to 10s
+- **Auto-formatter tool paths** — resolved via `which` first
+- **Protected-files-guard regex** — tightened `id_rsa` pattern
+- **Syntax-validator JSONC regex** — handles URLs containing `://`
+- **Skill-suggester keywords** — consolidated claude-agent-sdk phrases; added "compose" to docker
+- **redirect-builtin-agents.py fix** — `updatedInput` now preserves all original tool input fields (Claude Code replaces rather than merges)
+- **System prompt hardened** — added anti-fabrication rule, failure recovery strategy, and silent-violation guard to `execution_discipline` and `rule_precedence`
+
+### Removed
+
+- **setup-irie-claude.sh** — deleted (personal script, no longer invoked)
+- **output-style widget** — removed from ccstatusline (low value)
+
+### Documentation
+
+- **CLAUDE.md** — added keybindings.json, updated plugins list, fixed model name, documented VS Code conflicts, documented OVERWRITE_CONFIG, added agents/skills sections, added new scripts
+- **README.md** — fixed max output tokens, added keybindings section, added agents/skills, added project manager
+- **features/README.md** — full rewrite listing all features
+- **CHANGELOG.md** — squashed v1.6.0 + v1.6.1 into this entry
+
+---
+
 ## [v1.5.8] - 2026-02-06
 
 ### Changed

package/.devcontainer/CLAUDE.md

@@ -11,13 +11,15 @@ CodeForge devcontainer for AI-assisted development with Claude Code.
 │   ├── .env                 # Environment variables
 │   ├── config/              # Default configurations
 │   │   ├── settings.json    # Claude Code settings
+│   │   ├── keybindings.json # Claude Code keybindings
 │   │   └── main-system-prompt.md
 │   ├── features/            # Custom devcontainer features
 │   ├── plugins/             # Local plugin marketplace
 │   │   └── devs-marketplace/
 │   └── scripts/             # Setup scripts
 ├── .claude/                 # Runtime Claude config (created on first run)
-│   ├── settings.json        # Active settings
+│   ├── settings.json        # Active settings (OVERWRITE_CONFIG=true → rebuilt each start)
+│   ├── keybindings.json     # Active keybindings
 │   └── system-prompt.md     # Active system prompt
 └── .gh/                     # GitHub CLI config (persists across rebuilds)
     └── hosts.yml            # Authenticated hosts
@@ -30,8 +32,11 @@ CodeForge devcontainer for AI-assisted development with Claude Code.
 | `devcontainer.json` | Container definition: base image, features, mounts, environment |
 | `.env` | Environment variables controlling setup behavior |
 | `config/settings.json` | Claude Code defaults: model, tokens, permissions, plugins |
+| `config/keybindings.json` | Claude Code keybindings (empty by default — customizable) |
 | `config/main-system-prompt.md` | Default system prompt defining assistant behavior |
 
+> **Note**: `OVERWRITE_CONFIG=true` (default) means `.claude/settings.json` is overwritten on every container start. All persistent changes must go in `.devcontainer/config/settings.json`.
+
 ## Commands
 
 | Command | Purpose |
@@ -63,6 +68,8 @@ To test a feature locally, reference it in `devcontainer.json`:
 }
 ```
 
+> **Note**: The old `./features/claude-code` local feature is no longer used. Claude Code is installed via `ghcr.io/anthropics/devcontainer-features/claude-code:1` (Anthropic's official feature). The local `features/claude-code/` directory contains only a fallback config for the feature's install script.
+
 ## Setup Scripts
 
 Scripts in `./scripts/` run via `postStartCommand`:
@@ -70,17 +77,29 @@ Scripts in `./scripts/` run via `postStartCommand`:
 | Script | Purpose |
 |--------|---------|
 | `setup.sh` | Main orchestrator |
-| `setup-config.sh` | Copies config files to `/workspaces/.claude/` |
-| `setup-aliases.sh` | Creates `cc` shell function |
-| `setup-lsp.sh` | Installs LSP plugins from local marketplace |
-| `setup-plugins.sh` | Installs official Anthropic plugins |
+| `setup-config.sh` | Copies config files (settings, keybindings, system prompt) to `/workspaces/.claude/` |
+| `setup-aliases.sh` | Creates `cc`/`claude`/`ccraw` shell aliases |
+| `setup-plugins.sh` | Registers local marketplace + installs official Anthropic plugins |
+| `setup-update-claude.sh` | Background auto-update of Claude Code binary |
+| `setup-projects.sh` | Auto-detects projects for VS Code Project Manager |
+| `setup-symlink-claude.sh` | Symlinks ~/.claude for third-party tool compatibility |
 
 ## Installed Plugins
 
-Automatically installed on container start:
+Plugins are declared in `config/settings.json` under `enabledPlugins` and auto-activated on container start:
+
+### Official (Anthropic)
+- `frontend-design@claude-code-plugins` — UI/frontend design skill
 
-- `frontend-design@claude-plugins-official` (skill)
-- `codeforge-lsp@devs-marketplace` (LSP for Python + TypeScript/JavaScript)
+### Local Marketplace (devs-marketplace)
+- `codeforge-lsp@devs-marketplace` — LSP for Python + TypeScript/JavaScript
+- `ticket-workflow@devs-marketplace` — EARS-based ticket workflow with GitHub integration
+- `notify-hook@devs-marketplace` — Desktop notifications on completion
+- `dangerous-command-blocker@devs-marketplace` — Blocks destructive bash commands
+- `protected-files-guard@devs-marketplace` — Blocks edits to secrets/lock files
+- `auto-formatter@devs-marketplace` — Auto-formats Python/Go/JS/TS on stop
+- `auto-linter@devs-marketplace` — Auto-lints Python files via Pyright
+- `code-directive@devs-marketplace` — 17 custom agents, 16 skills, syntax validation, skill suggestions, agent redirect hook
 
 ### Local Marketplace
 
@@ -91,12 +110,41 @@ plugins/devs-marketplace/
 ├── .claude-plugin/
 │   └── marketplace.json      # Marketplace manifest
 └── plugins/
-    └── codeforge-lsp/        # Combined LSP plugin
-        └── .claude-plugin/
-            └── plugin.json
+    ├── codeforge-lsp/        # Combined LSP plugin
+    ├── ticket-workflow/      # EARS ticket workflow
+    ├── auto-formatter/       # Batch formatter (Stop hook)
+    ├── auto-linter/          # Pyright linter
+    ├── code-directive/       # Agents, skills + hooks
+    └── ...
 ```
 
-This gives full control over LSP configuration without external dependencies.
+## Agents & Skills
+
+The `code-directive` plugin includes 17 custom agent definitions and 16 coding reference skills.
+
+**Agents** (`plugins/devs-marketplace/plugins/code-directive/agents/`):
+architect, bash-exec, claude-guide, debug-logs, dependency-analyst, doc-writer, explorer, generalist, git-archaeologist, migrator, perf-profiler, refactorer, researcher, security-auditor, spec-writer, statusline-config, test-writer
+
+The `redirect-builtin-agents.py` hook (PreToolUse/Task) transparently swaps built-in agent types to these custom agents (e.g., Explore→explorer, Plan→architect).
+
+**Skills** (`plugins/devs-marketplace/plugins/code-directive/skills/`):
+claude-agent-sdk, claude-code-headless, debugging, docker, docker-py, fastapi, git-forensics, performance-profiling, pydantic-ai, refactoring-patterns, security-checklist, skill-building, specification-writing, sqlite, svelte5, testing
+
+## VS Code Keybinding Conflicts
+
+Claude Code runs inside VS Code's integrated terminal. VS Code intercepts some shortcuts before they reach the terminal:
+
+| Shortcut | VS Code Action | Claude Code Action |
+|----------|---------------|-------------------|
+| `Ctrl+G` | Go to Line | `chat:externalEditor` |
+| `Ctrl+S` | Save File | `chat:stash` |
+| `Ctrl+T` | Open Symbol | `app:toggleTodos` |
+| `Ctrl+O` | Open File | `app:toggleTranscript` |
+| `Ctrl+B` | Toggle Sidebar | `task:background` |
+| `Ctrl+P` | Quick Open | `chat:modelPicker` |
+| `Ctrl+R` | Open Recent | `history:search` |
+
+`Ctrl+P` and `Ctrl+F` are configured to pass through to the terminal via `terminal.integrated.commandsToSkipShell` in `devcontainer.json`. For other conflicts, use Meta (Alt) variants or customize via `config/keybindings.json`.
 
 ## Environment Variables
 
@@ -107,11 +155,13 @@ Key environment variables set in the container:
 | `WORKSPACE_ROOT` | `/workspaces` |
 | `CLAUDE_CONFIG_DIR` | `/workspaces/.claude` |
 | `GH_CONFIG_DIR` | `/workspaces/.gh` |
-| `ANTHROPIC_MODEL` | `claude-opus-4-5-20251101` |
+| `ANTHROPIC_MODEL` | `claude-opus-4-6` |
+| `TMPDIR` | `/workspaces/.tmp` |
 
 ## Modifying Behavior
 
 1. **Change default model**: Edit `config/settings.json`, update `"model"` field
 2. **Change system prompt**: Edit `config/main-system-prompt.md`
-3. **Add features**: Add to `"features"` in `devcontainer.json`
-4. **Disable auto-setup**: Set variables to `false` in `.env`
+3. **Change keybindings**: Edit `config/keybindings.json`
+4. **Add features**: Add to `"features"` in `devcontainer.json`
+5. **Disable auto-setup**: Set variables to `false` in `.env`

package/.devcontainer/README.md

@@ -143,7 +143,8 @@ GitHub CLI credentials are automatically persisted across container rebuilds. Th
 | `claude` | Claude Code CLI |
 | `cc` | Wrapper with auto-configuration |
 | `ccusage` | Token usage analyzer |
-| `ccstatusline` | Status line for sessions |
+| `ccburn` | Visual token burn rate tracker with pace indicators |
+| `ccstatusline` | 6-line powerline status display (v1.1.0) |
 | `claude-monitor` | Real-time usage tracking |
 
 ## Using Claude Code
@@ -188,7 +189,24 @@ Default settings are in `.devcontainer/config/settings.json`. These are copied t
 Key defaults:
 - **Model**: Claude Opus 4-6
 - **Default mode**: Plan (prompts before executing)
-- **Max output tokens**: 128,000
+- **Max output tokens**: 64,000
+
+### Keybindings
+
+Default keybindings are in `.devcontainer/config/keybindings.json` (empty by default — Claude Code defaults apply). Customize by adding entries to the `bindings` array.
+
+**VS Code Terminal Passthrough**: `Ctrl+P` and `Ctrl+F` are configured to pass through to the terminal (via `terminal.integrated.commandsToSkipShell`) so Claude Code receives them. Other VS Code shortcuts that conflict with Claude Code:
+
+| Shortcut | VS Code Action | Claude Code Action |
+|----------|---------------|-------------------|
+| `Ctrl+G` | Go to Line | `chat:externalEditor` |
+| `Ctrl+S` | Save File | `chat:stash` |
+| `Ctrl+T` | Open Symbol | `app:toggleTodos` |
+| `Ctrl+O` | Open File | `app:toggleTranscript` |
+| `Ctrl+B` | Toggle Sidebar | `task:background` |
+| `Ctrl+R` | Open Recent | `history:search` |
+
+For conflicting shortcuts, use Meta (Alt) variants or add custom keybindings.
 
 ### System Prompt
 
@@ -200,14 +218,57 @@ CodeForge includes several custom devcontainer features:
 
 | Feature | Description |
 |---------|-------------|
+| `tmux` | Terminal multiplexer with Catppuccin theme for Agent Teams |
+| `agent-browser` | Headless browser automation for AI agents |
 | `claude-monitor` | Real-time token usage monitoring with ML predictions |
 | `ccusage` | Usage analytics CLI |
-| `ccstatusline` | Compact powerline status display |
+| `ccburn` | Visual token burn rate tracker with pace indicators |
+| `ccstatusline` | 6-line powerline status display (v1.1.0) |
 | `ast-grep` | Structural code search using AST patterns |
 | `tree-sitter` | Parser with JS/TS/Python grammars |
 | `lsp-servers` | Pyright and TypeScript language servers |
-| `tmux` | Terminal multiplexer with Catppuccin theme for Agent Teams |
 | `biome` | Fast JS/TS/JSON/CSS formatter (global install) |
+| `notify-hook` | Desktop notifications on Claude completion |
+| `mcp-qdrant` | Qdrant vector database MCP server (optional) |
+| `mcp-reasoner` | Enhanced AI reasoning MCP server (optional) |
+
+## Agents & Skills
+
+The `code-directive` plugin includes specialized agent definitions and coding reference skills.
+
+### Custom Agents (17)
+
+Agent definitions in `plugins/devs-marketplace/plugins/code-directive/agents/` provide enhanced behavior when spawned via the `Task` tool. The `redirect-builtin-agents.py` hook transparently swaps built-in agent types to these custom agents.
+
+| Agent | Purpose |
+|-------|---------|
+| `architect` | System design and implementation planning |
+| `bash-exec` | Command execution specialist |
+| `claude-guide` | Claude Code feature guidance |
+| `debug-logs` | Log analysis and error diagnosis |
+| `dependency-analyst` | Dependency analysis and upgrades |
+| `doc-writer` | Documentation authoring |
+| `explorer` | Fast codebase search and navigation |
+| `generalist` | General-purpose multi-step tasks |
+| `git-archaeologist` | Git history forensics |
+| `migrator` | Code migration and upgrades |
+| `perf-profiler` | Performance profiling |
+| `refactorer` | Code refactoring with regression checks |
+| `researcher` | Research and information gathering |
+| `security-auditor` | Security vulnerability analysis |
+| `spec-writer` | Specification and requirements authoring |
+| `statusline-config` | ccstatusline configuration |
+| `test-writer` | Test authoring with pass verification |
+
+### Skills (16)
+
+Skills in `plugins/devs-marketplace/plugins/code-directive/skills/` provide domain-specific coding references:
+
+`claude-agent-sdk` · `claude-code-headless` · `debugging` · `docker` · `docker-py` · `fastapi` · `git-forensics` · `performance-profiling` · `pydantic-ai` · `refactoring-patterns` · `security-checklist` · `skill-building` · `specification-writing` · `sqlite` · `svelte5` · `testing`
+
+## Project Manager
+
+The `setup-projects.sh` script auto-detects projects under `/workspaces/` and maintains a `projects.json` file for the [Project Manager](https://marketplace.visualstudio.com/items?itemName=alefragnani.project-manager) VS Code extension. It watches for new projects via `inotifywait` and updates the project list automatically.
 
 ## Essential Gotchas
 

package/.devcontainer/config/keybindings.json

@@ -0,0 +1,5 @@
+{
+  "$schema": "https://www.schemastore.org/claude-code-keybindings.json",
+  "$docs": "https://code.claude.com/docs/en/keybindings",
+  "bindings": []
+}

package/.devcontainer/config/main-system-prompt.md

@@ -25,7 +25,7 @@ When in <normal_mode>:
 8. <response_guidelines>
 
 If rules conflict, follow the highest-priority rule for the active mode
-and explicitly note the conflict.
+and explicitly note the conflict. Never silently violate a higher-priority rule.
 </rule_precedence>
 
 <operating_modes>
@@ -121,7 +121,13 @@ Main thread:
 Subagents (via `Task`):
 - Information gathering only
 - Report findings; never decide or modify
-- Types: `Explore` (fast search), `Plan` (design), `general-purpose` (complex), `Bash` (commands)
+- Core types (auto-redirected to enhanced custom agents):
+  - `Explore` → `explorer` (fast codebase search, haiku, read-only)
+  - `Plan` → `architect` (implementation planning, opus, read-only)
+  - `general-purpose` → `generalist` (multi-step tasks, inherit model)
+  - `Bash` → `bash-exec` (command execution, sonnet)
+  - `claude-code-guide` → `claude-guide` (Claude Code/SDK/API help, haiku)
+  - `statusline-setup` → `statusline-config` (status line setup, sonnet)
 
 Agent Teams (when enabled):
 - CRITICAL: Limit to 3-5 active teammates maximum based on task complexity
@@ -145,6 +151,55 @@ Failure handling:
 - Surface errors clearly; never hide failures
 </orchestration>
 
+<specialist_agents>
+Specialist agents are available as teammates via the Task tool. Prefer
+delegating to a specialist over doing the work yourself when the task
+matches their domain.
+
+Agents:
+- researcher — codebase & web research (sonnet, read-only)
+- test-writer — writes test suites (opus, auto-verify)
+- refactorer — safe code transformations (opus, tests after every edit)
+- security-auditor — OWASP audit & secrets scan (sonnet, read-only)
+- doc-writer — README, API docs, docstrings (opus)
+- migrator — framework upgrades & version bumps (opus)
+- git-archaeologist — git history investigation (haiku, read-only)
+- dependency-analyst — outdated/vulnerable deps (haiku, read-only)
+- spec-writer — EARS requirements & acceptance criteria (opus, read-only)
+- perf-profiler — profiling & benchmarks (sonnet, read-only)
+- debug-logs — log analysis & diagnostics (sonnet, read-only)
+
+Skills (auto-suggested, also loadable via Skill tool):
+- fastapi, sqlite, svelte5, docker, docker-py, pydantic-ai
+- testing, debugging, claude-code-headless, claude-agent-sdk
+- skill-building, refactoring-patterns, security-checklist
+- git-forensics, specification-writing, performance-profiling
+
+Built-in agent redirect:
+All 6 built-in agent types (Explore, Plan, general-purpose, Bash,
+claude-code-guide, statusline-setup) are automatically redirected to
+enhanced custom agents via a PreToolUse hook. You can use either the
+built-in name or the custom name — the redirect is transparent.
+
+Team construction:
+When building agent teams, prefer custom agents over generic
+`generalist` teammates when the task aligns with a specialist's
+domain. Custom agents carry frontloaded skills, safety hooks, and
+tailored instructions that make them more effective and safer than
+a generalist agent doing the same work.
+
+Example team compositions:
+- Feature build: researcher (investigate) + test-writer (tests) + doc-writer (docs)
+- Security hardening: security-auditor (find issues) + dependency-analyst (deps)
+- Codebase cleanup: refactorer (transform) + test-writer (coverage gaps)
+- Migration project: researcher (research guides) + migrator (execute)
+- Performance work: perf-profiler (measure) + refactorer (optimize)
+
+When a user's request clearly falls within a specialist's domain,
+suggest delegation. Do not force it — the user may prefer to work
+directly.
+</specialist_agents>
+
 <structural_search>
 Prefer structural tools over text search when syntax matters:
 
@@ -258,6 +313,7 @@ Verify before assuming:
   or approach — ASK. Do not pick a default.
 - Do not assume file paths — read the filesystem to confirm.
 - Do not assume platform capabilities — research first.
+- Never fabricate file paths, API signatures, tool behavior, or external facts. Verify or ask.
 
 Read before writing:
 - Before creating or modifying any file, read the target directory and
@@ -286,6 +342,11 @@ No silent deviations:
   before doing something different.
 - Never silently substitute an easier approach.
 - Never silently skip a step because it seems hard or uncertain.
+
+When an approach fails:
+- Diagnose the cause before retrying.
+- Try an alternative strategy; do not repeat the failed path.
+- Surface the failure and revised approach to the user.
 </execution_discipline>
 
 <code_directives>

package/.devcontainer/config/settings.json

@@ -9,19 +9,30 @@
     "ANTHROPIC_DEFAULT_HAIKU_MODEL": "claude-haiku-4-5-20251001",
     "BASH_DEFAULT_TIMEOUT_MS": "240000",
     "BASH_MAX_TIMEOUT_MS": "600000",
-    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "true",
-    "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "128000",
+    "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "64000",
     "MAX_MCP_OUTPUT_TOKENS": "10000",
     "MAX_THINKING_TOKENS": "63999",
     "MCP_TIMEOUT": "120000",
     "MCP_TOOL_TIMEOUT": "30000",
-    "CLAUDE_AUTOCOMPACT_PCT_OVERRIDE": 95,
+    "CLAUDE_AUTOCOMPACT_PCT_OVERRIDE": "95",
     "CLAUDE_CODE_SHELL": "zsh",
-    "FORCE_AUTOUPDATE_PLUGINS": true,
+    "FORCE_AUTOUPDATE_PLUGINS": "1",
+
     "ENABLE_TOOL_SEARCH": "auto:5",
     "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1",
     "CLAUDE_CODE_EFFORT_LEVEL": "high",
-    "CLAUDE_CODE_ENABLE_TASKS": "true"
+    "CLAUDE_CODE_ENABLE_TASKS": "true",
+    "CLAUDE_CODE_DISABLE_AUTO_MEMORY": "0",
+    "ENABLE_CLAUDE_CODE_SM_COMPACT": "1",
+    "CLAUDE_CODE_FORCE_GLOBAL_CACHE": "1",
+    "CLAUDE_CODE_PLAN_MODE_INTERVIEW_PHASE": "true",
+    "CLAUDE_CODE_PLAN_V2_AGENT_COUNT": "3",
+    "CLAUDE_CODE_PLAN_MODE_REQUIRED": "true",
+
+    "CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY": "5",
+    "CLAUDE_CODE_MAX_RETRIES": "1",
+    "BASH_MAX_OUTPUT_LENGTH": "15000",
+    "TASK_MAX_OUTPUT_LENGTH": "64000"
   },
   "teammateMode": "auto",
   "effortLevel": "high",
@@ -45,7 +56,15 @@
     "command": "/usr/local/bin/ccstatusline-wrapper"
   },
   "enabledPlugins": {
-    "frontend-design@claude-code-plugins": true
+    "frontend-design@claude-code-plugins": true,
+    "codeforge-lsp@devs-marketplace": true,
+    "ticket-workflow@devs-marketplace": true,
+    "notify-hook@devs-marketplace": true,
+    "dangerous-command-blocker@devs-marketplace": true,
+    "protected-files-guard@devs-marketplace": true,
+    "auto-formatter@devs-marketplace": true,
+    "auto-linter@devs-marketplace": true,
+    "code-directive@devs-marketplace": true
   },
   "autoUpdatesChannel": "latest"
 }

package/.devcontainer/devcontainer.json

@@ -24,6 +24,7 @@
     "./features/agent-browser",
     "./features/claude-monitor",
     "./features/ccusage",
+    "./features/ccburn",
     "./features/ccstatusline",
     "./features/ast-grep",
     "./features/tree-sitter",
@@ -56,6 +57,11 @@
       "installer": "uv",
       "username": "automatic"
     },
+    "./features/ccburn": {
+      "version": "latest",
+      "shells": "both",
+      "username": "automatic"
+    },
     "./features/ccstatusline": {
       "username": "automatic"
     },
@@ -89,14 +95,24 @@
         },
         "terminal.integrated.defaultProfile.linux": "bash",
         "terminal.integrated.enableBell": true,
+        "terminal.integrated.commandsToSkipShell": [
+          "-workbench.action.quickOpen",
+          "-workbench.action.terminal.focusFind"
+        ],
         "remote.extensionKind": {
           "wenbopan.vscode-terminal-osc-notifier": ["ui"]
-        }
+        },
+        "projectManager.git.baseFolders": ["/workspaces"],
+        "projectManager.git.maxDepthRecursion": 1,
+        "projectManager.showProjectNameInStatusBar": true,
+        "projectManager.openInNewWindowWhenClickingInStatusBar": false,
+        "projectManager.projectsLocation": "/workspaces/.config/project-manager"
       },
       "extensions": [
         "wenbopan.vscode-terminal-osc-notifier",
         "GitHub.vscode-github-actions",
-        "fabiospampinato.vscode-todo-plus"
+        "fabiospampinato.vscode-todo-plus",
+        "alefragnani.project-manager"
       ]
     }
   },

package/.devcontainer/features/README.md

@@ -4,11 +4,25 @@ This directory contains DevContainer Features for AI coding agent environments.
 
 ## Available Features
 
-### mcp-qdrant
-Vector database integration via Qdrant MCP server. Enables semantic search, embeddings management, and RAG workflows.
-
-**Status**: ✅ Complete
-**Documentation**: [mcp-qdrant/README.md](./mcp-qdrant/README.md)
+| Feature | Description | Status |
+|---------|-------------|--------|
+| `tmux` | Terminal multiplexer with Catppuccin theme for Agent Teams | ✅ |
+| `agent-browser` | Headless browser automation for AI agents | ✅ |
+| `claude-monitor` | Real-time token usage monitoring | ✅ |
+| `ccusage` | Token usage analytics CLI | ✅ |
+| `ccburn` | Visual token burn rate tracker with pace indicators | ✅ New |
+| `ccstatusline` | 6-line powerline status display (v1.1.0) | ✅ |
+| `ast-grep` | Structural code search using AST patterns | ✅ |
+| `tree-sitter` | Parser with JS/TS/Python grammars | ✅ |
+| `lsp-servers` | Pyright and TypeScript language servers | ✅ |
+| `biome` | Fast JS/TS/JSON/CSS formatter | ✅ |
+| `notify-hook` | Desktop notifications on Claude completion | ✅ |
+| `splitrail` | Git worktree management for parallel branches | ✅ |
+| `mcp-qdrant` | Qdrant vector database MCP server | ✅ (optional) |
+| `mcp-reasoner` | Enhanced AI reasoning MCP server | ✅ (optional) |
+| `claude-code` | Fallback config for Anthropic's official Claude Code feature | ✅ (config only) |
+
+> **Note**: Claude Code itself is installed via `ghcr.io/anthropics/devcontainer-features/claude-code:1` (Anthropic's official feature). The local `claude-code/` directory provides only fallback configuration.
 
 ## Feature Structure
 
@@ -90,9 +104,9 @@ CodeForge modules are being converted to DevContainer Features:
 | Old Module | New Feature | Status |
 |-----------|-------------|---------|
 | mcp_qdrant | mcp-qdrant | ✅ Complete |
+| mcp_reasoner | mcp-reasoner | ✅ Complete |
 | mcp_browser | mcp-browser | 🔜 Planned |
 | mcp_sourcerer | mcp-sourcerer | 🔜 Planned |
-| mcp_reasoner | mcp-reasoner | 🔜 Planned |
 | mcp_obsidian | mcp-obsidian | 🔜 Planned |
 | mcp_code_runner | mcp-code-runner | 🔜 Planned |
 
@@ -110,4 +124,4 @@ Features are part of the CodeForge project. See main repository for contribution
 ---
 
 **Status**: Active Development
-**Last Updated**: 2025-11-11
+**Last Updated**: 2026-02-08

package/.devcontainer/features/ccburn/README.md

@@ -0,0 +1,60 @@
+# ccburn - Visual Token Burn Rate Tracker
+
+Real-time burn-up charts and pace tracking for Claude Code usage limits.
+
+## Quick Start
+
+```bash
+# Full TUI dashboard (auto-detects active limit)
+ccburn
+
+# 5-hour rolling session limit
+ccburn session
+
+# 7-day weekly limit
+ccburn weekly
+
+# Single-line output for status bars
+ccburn --compact --once
+
+# JSON output for scripting
+ccburn --json --once
+```
+
+## Features
+
+- **Burn-up charts** — Live-updating terminal graphics showing usage over time
+- **Pace indicators** — 🧊 Cool (behind pace) | 🔥 On pace | 🚨 Too hot (ahead)
+- **Burn rate** — %/hour via linear regression, with trend classification
+- **Predictions** — Time-to-100% estimates and projection lines
+- **Multiple limits** — Session (5h), Weekly (7d), Weekly Sonnet, Monthly
+- **SQLite history** — 7-day retention at `~/.ccburn/history.db`
+- **Compact mode** — Single-line output for embedding in statuslines
+
+## Display Modes
+
+| Mode | Flag | Description |
+|------|------|-------------|
+| Full TUI | (default) | Interactive dashboard with charts and gauges |
+| Compact | `--compact` | Single line: `Session: 🧊 45% (2h14m) \| Weekly: 🔥 12%` |
+| JSON | `--json` | Structured output with all metrics |
+| Once | `--once` | Single snapshot, no live updates |
+
+## Statusline Integration
+
+This feature installs a wrapper at `/usr/local/bin/ccburn-statusline` for use with
+ccstatusline's `custom-command` widget. The wrapper handles:
+
+- Missing OAuth credentials (shows "awaiting auth")
+- First-run npx caching (shows "unavailable" if download fails)
+- API errors (graceful fallback messages)
+
+## Requirements
+
+- Node.js (via devcontainer node feature)
+- Claude Code OAuth credentials at `~/.claude/.credentials.json`
+
+## Documentation
+
+- [GitHub Repository](https://github.com/JuanjoFuchs/ccburn)
+- [ccburn Specification](https://github.com/JuanjoFuchs/ccburn/blob/main/docs/ccburn-spec.md)

package/.devcontainer/features/ccburn/devcontainer-feature.json

@@ -0,0 +1,38 @@
+{
+  "id": "ccburn",
+  "version": "1.0.0",
+  "name": "ccburn - Visual Token Burn Rate Tracker",
+  "description": "Installs ccburn CLI for real-time burn-up charts and pace tracking of Claude Code usage limits",
+  "maintainer": "AnExiledDev",
+  "documentationURL": "https://github.com/JuanjoFuchs/ccburn",
+  "options": {
+    "version": {
+      "type": "string",
+      "description": "ccburn version to use (e.g., 'latest', '0.4.2')",
+      "default": "latest"
+    },
+    "shells": {
+      "type": "string",
+      "description": "Which shells to configure (bash, zsh, or both)",
+      "default": "both",
+      "enum": [
+        "bash",
+        "zsh",
+        "both"
+      ],
+      "proposals": [
+        "bash",
+        "zsh",
+        "both"
+      ]
+    },
+    "username": {
+      "type": "string",
+      "description": "Container user to install for",
+      "default": "automatic"
+    }
+  },
+  "installsAfter": [
+    "ghcr.io/devcontainers/features/node:1"
+  ]
+}