loki-mode 5.49.1 → 5.49.2

package/README.md

@@ -11,7 +11,7 @@
 [![Agent Types](https://img.shields.io/badge/Agent%20Types-41-blue)]()
 [![Benchmarks](https://img.shields.io/badge/Benchmarks-Infrastructure%20Ready-blue)](benchmarks/)
 
-**Current Version: v5.49.0**
+**Current Version: v5.49.2**
 
 **[Autonomi](https://www.autonomi.dev/)** | **[Documentation](https://www.autonomi.dev/docs)** | **[GitHub](https://github.com/asklokesh/loki-mode)**
 
@@ -25,7 +25,7 @@
 
 [![asciicast](https://asciinema.org/a/AjjnjzOeKLYItp6s.svg)](https://asciinema.org/a/AjjnjzOeKLYItp6s)
 
-*Click to watch Loki Mode v5.42 -- CLI commands, dashboard, 8 parallel agents, 7-gate quality, Completion Council, memory system*
+*Click to watch Loki Mode v5.42 -- CLI commands, dashboard, 8 parallel agents, 9-gate quality, Completion Council, memory system*
 
 ---
 
@@ -39,98 +39,38 @@
 
 ---
 
-## Usage
-
-### Option 1: npm (Recommended)
+## Installation
 
 ```bash
-npm install -g loki-mode
-loki start ./my-prd.md
+git clone https://github.com/asklokesh/loki-mode.git ~/.claude/skills/loki-mode
 ```
 
-### Option 2: Claude Code Skill
+That's it. Claude Code auto-discovers skills in `~/.claude/skills/`.
+
+### Use It
 
 ```bash
-git clone https://github.com/asklokesh/loki-mode.git ~/.claude/skills/loki-mode
 claude --dangerously-skip-permissions
-# Then say: Loki Mode with PRD at ./my-prd.md
+# Then say: "Loki Mode with PRD at ./my-prd.md"
 ```
 
-### Option 3: GitHub Action
+### Update
 
-Add automated AI code review to your pull requests:
-
-```yaml
-# .github/workflows/loki-review.yml
-name: Loki Code Review
-
-on:
-  pull_request:
-    types: [opened, synchronize]
-
-permissions:
-  contents: read
-  pull-requests: write
-
-jobs:
-  review:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - uses: asklokesh/loki-mode@v5
-        with:
-          github_token: ${{ secrets.GITHUB_TOKEN }}
-          mode: review          # review, fix, or test
-          provider: claude      # claude, codex, or gemini
-          max_iterations: 3     # sets LOKI_MAX_ITERATIONS env var
-          budget_limit: '5.00'  # max cost in USD (maps to --budget flag)
-        env:
-          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-```
-
-**Prerequisites:**
-- An API key for your chosen provider (set as a repository secret):
-  - Claude: `ANTHROPIC_API_KEY`
-  - Codex: `OPENAI_API_KEY`
-  - Gemini: `GOOGLE_API_KEY`
-- The action automatically installs `loki-mode` and `@anthropic-ai/claude-code` (for the Claude provider)
-
-**Action Inputs:**
-
-| Input | Default | Description |
-|-------|---------|-------------|
-| `mode` | `review` | `review`, `fix`, or `test` |
-| `provider` | `claude` | `claude`, `codex`, or `gemini` |
-| `budget_limit` | `5.00` | Max cost in USD (maps to `--budget` CLI flag) |
-| `budget` | | Alias for `budget_limit` |
-| `max_iterations` | `3` | Sets `LOKI_MAX_ITERATIONS` env var |
-| `github_token` | (required) | GitHub token for PR comments |
-| `prd_file` | | Path to PRD file relative to repo root |
-| `auto_confirm` | `true` | Skip confirmation prompts (always true in CI) |
-| `install_claude` | `true` | Auto-install Claude Code CLI if not present |
-| `node_version` | `20` | Node.js version |
-
-**Using with a PRD file (fix/test modes):**
-
-```yaml
-- uses: asklokesh/loki-mode@v5
-  with:
-    mode: fix
-    prd_file: 'docs/my-prd.md'
-    github_token: ${{ secrets.GITHUB_TOKEN }}
-  env:
-    ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+```bash
+cd ~/.claude/skills/loki-mode && git pull
 ```
 
-**Modes:**
+### Troubleshooting
+
+| Problem | Fix |
+|---------|-----|
+| `SKILL.md` not found | Verify: `ls ~/.claude/skills/loki-mode/SKILL.md` |
+| Claude doesn't recognize "Loki Mode" | Restart Claude Code after cloning |
+| Permission denied on clone | Check SSH keys or use HTTPS URL above |
 
-| Mode | Description |
-|------|-------------|
-| `review` | Analyze PR diff, post structured review as PR comment |
-| `fix` | Automatically fix issues found in the codebase |
-| `test` | Run autonomous test generation and validation |
+### Other Installation Methods
 
-Also available via **Homebrew**, **Docker**, **VS Code Extension**, and **direct shell script**. See the [Installation Guide](docs/INSTALLATION.md) for all 7 installation methods and detailed instructions.
+Also available via **npm**, **Homebrew**, **Docker**, **GitHub Action**, and **VS Code Extension**. See [docs/alternative-installations.md](docs/alternative-installations.md) for details and limitations of each method.
 
 ### Multi-Provider Support (v5.0.0)
 
@@ -188,6 +128,30 @@ PRD → Research → Architecture → Development → Testing → Deployment →
 
 ---
 
+## Current Limitations
+
+Loki Mode is powerful but not magic. Be aware of these honest limitations:
+
+| Area | What Works | What Doesn't (Yet) |
+|------|-----------|---------------------|
+| **Code Generation** | Generates full-stack applications from PRDs | Complex domain logic may need human review and correction |
+| **Deployment** | Generates deployment configs and scripts | Does not have cloud credentials -- human must provide and authorize |
+| **Testing** | 9 automated quality gates, blind review | Test quality depends on AI-generated assertions; mutation testing is heuristic |
+| **Business Ops** | Generates marketing copy, legal templates | Does not actually send emails, file legal documents, or process payments |
+| **Multi-Provider** | Claude (full), Codex (degraded), Gemini (degraded) | Codex and Gemini lack parallel agents and Task tool -- sequential only |
+| **Memory System** | Episodic, semantic, procedural memory tiers | Vector search requires optional `sentence-transformers` dependency |
+| **Enterprise Security** | TLS, OIDC, RBAC, audit trail, SIEM configs | Self-signed certs only; production deployments need real certificates |
+| **Dashboard** | Real-time status, task queue, agent monitoring | Single-machine only; no multi-node dashboard clustering |
+| **Benchmarks** | HumanEval 98.78%, SWE-bench 299/300 patches | Self-reported; SWE-bench counts patch generation, not verified resolution |
+
+**What "autonomous" means in practice:**
+- Loki Mode runs without prompting between RARV cycles
+- It does NOT have access to your cloud accounts, payment systems, or external services unless you provide credentials
+- Human oversight is expected for: deployment credentials, domain setup, API keys, and critical business decisions
+- The system is as good as the underlying AI model -- it can make mistakes, especially on novel or complex problems
+
+---
+
 ## Why Loki Mode?
 
 ### **How It Works**
@@ -196,7 +160,7 @@ PRD → Research → Architecture → Development → Testing → Deployment →
 |----------------|---------------------|
 | **Single agent** writes code linearly | **Multiple agents** work in parallel across engineering, ops, business, data, product, and growth |
 | **Manual deployment** required | **Autonomous deployment** to AWS, GCP, Azure, Vercel, Railway with blue-green and canary strategies |
-| **No testing** or basic unit tests | **7 automated quality gates**: input/output guardrails, static analysis, blind review, anti-sycophancy, severity blocking, test coverage |
+| **No testing** or basic unit tests | **9 automated quality gates**: input/output guardrails, static analysis, blind review, anti-sycophancy, severity blocking, test coverage, mock detection, mutation detection |
 | **Code only** - you handle the rest | **Full business operations**: marketing, sales, legal, HR, finance, investor relations |
 | **Stops on errors** | **Self-healing**: circuit breakers, dead letter queues, exponential backoff, automatic recovery |
 | **No visibility** into progress | **Real-time dashboard** with agent monitoring, task queues, and live status updates |
@@ -234,7 +198,7 @@ PRD → Research → Architecture → Development → Testing → Deployment →
 | **OpenClaw Bridge (v5.38.0)** | Multi-agent coordination protocol | [OpenClaw Integration](docs/openclaw-integration.md) |
 | **41 Agent Types** | Engineering, Ops, Business, Data, Product, Growth, Orchestration | [Agent Definitions](references/agent-types.md) |
 | **RARV Cycle** | Reason-Act-Reflect-Verify workflow | [Core Workflow](references/core-workflow.md) |
-| **Quality Gates** | 7-gate system: guardrails, static analysis, blind review, anti-sycophancy, severity blocking, test coverage | [Quality Control](references/quality-control.md) |
+| **Quality Gates** | 9-gate system: guardrails, static analysis, blind review, anti-sycophancy, severity blocking, test coverage, mock detection, mutation detection | [Quality Control](references/quality-control.md) |
 | **Memory System (v5.15.0)** | Complete 3-tier memory with progressive disclosure | [Memory Architecture](references/memory-system.md) |
 | **Parallel Workflows** | Git worktree-based parallelism | [Parallel Workflows](skills/parallel-workflows.md) |
 | **GitHub Integration** | Issue import, PR creation, status sync | [GitHub Integration](skills/github-integration.md) |
@@ -661,7 +625,7 @@ references/                    # Deep documentation (23KB+ files)
 | **2. Architecture** | Tech stack selection with self-reflection |
 | **3. Infrastructure** | Provision cloud, CI/CD, monitoring |
 | **4. Development** | Implement with TDD, parallel code review |
-| **5. QA** | 7 quality gates, security audit, load testing |
+| **5. QA** | 9 quality gates, security audit, load testing |
 | **6. Deployment** | Blue-green deploy, auto-rollback on errors |
 | **7. Business** | Marketing, sales, legal, support setup |
 | **8. Growth** | Continuous optimization, A/B testing, feedback loops |

package/SKILL.md

@@ -3,7 +3,7 @@ name: loki-mode
 description: Multi-agent autonomous startup system. Triggers on "Loki Mode". Takes PRD to deployed product with minimal human intervention. Requires --dangerously-skip-permissions flag.
 ---
 
-# Loki Mode v5.49.1
+# Loki Mode v5.49.2
 
 **You are an autonomous agent. You make decisions. You do not ask questions. You do not stop.**
 
@@ -263,4 +263,4 @@ The following features are documented in skill modules but not yet fully automat
 | Quality gates 3-reviewer system | Implemented (v5.35.0) | 5 specialist reviewers in `skills/quality-gates.md`; execution in run.sh |
 | Benchmarks (HumanEval, SWE-bench) | Infrastructure only | Runner scripts and datasets exist in `benchmarks/`; no published results |
 
-**v5.49.1 | [Autonomi](https://www.autonomi.dev/) flagship product | ~260 lines core**
+**v5.49.2 | [Autonomi](https://www.autonomi.dev/) flagship product | ~260 lines core**

package/VERSION

@@ -1 +1 @@
-5.49.1
+5.49.2

package/autonomy/CONSTITUTION.md

@@ -155,7 +155,7 @@ SKILL.md (~190 lines)         # Always loaded: RARV cycle, autonomy rules
 skills/
   00-index.md                  # Module routing table
   model-selection.md           # Task tool, parallelization
-  quality-gates.md             # 7-gate system, anti-sycophancy
+  quality-gates.md             # 9-gate system, anti-sycophancy
   testing.md                   # Playwright, E2E, property-based
   production.md                # CI/CD, batch processing
   agents.md                    # 41 agent types, A2A patterns
@@ -196,7 +196,7 @@ Main Worktree (orchestrator)
 
 ---
 
-## Quality Gates (7-Gate System)
+## Quality Gates (9-Gate System)
 
 ### Gate 1: Static Analysis
 ```yaml

package/dashboard/__init__.py

@@ -7,7 +7,7 @@ Modules:
     control: Session control API (start/stop/pause/resume)
 """
 
-__version__ = "5.49.1"
+__version__ = "5.49.2"
 
 # Expose the control app for easy import
 try:

package/dashboard/server.py

@@ -3151,23 +3151,112 @@ async def get_github_sync_log(
 # =============================================================================
 
 
+def _resolve_process_state(pid: Optional[int], last_status: str = "",
+                           started: str = "", heartbeat: str = "",
+                           stale_threshold: int = 30) -> dict[str, Any]:
+    """Resolve process state with honest labels.
+
+    States:
+      RUNNING   - PID alive AND heartbeat < stale_threshold seconds
+      STALE     - PID alive BUT no heartbeat update in > stale_threshold seconds
+      COMPLETED - last_status marked done/completed and PID exited
+      FAILED    - last_status marked failed OR PID exited non-zero
+      CRASHED   - PID dead BUT last_status was 'running'
+      UNKNOWN   - No PID, no status, or conflicting data
+
+    Returns dict with: state, pid_alive, started, last_heartbeat, duration_seconds
+    """
+    now = datetime.now(timezone.utc)
+    pid_alive = False
+    if pid is not None:
+        try:
+            os.kill(pid, 0)
+            pid_alive = True
+        except (OSError, ValueError, TypeError):
+            pass
+
+    # Parse timestamps
+    started_dt = None
+    heartbeat_dt = None
+    if started:
+        try:
+            started_dt = datetime.fromisoformat(started.replace("Z", "+00:00"))
+            if started_dt.tzinfo is None:
+                started_dt = started_dt.replace(tzinfo=timezone.utc)
+        except (ValueError, AttributeError):
+            pass
+    if heartbeat:
+        try:
+            heartbeat_dt = datetime.fromisoformat(heartbeat.replace("Z", "+00:00"))
+            if heartbeat_dt.tzinfo is None:
+                heartbeat_dt = heartbeat_dt.replace(tzinfo=timezone.utc)
+        except (ValueError, AttributeError):
+            pass
+
+    # Calculate duration
+    duration_seconds = None
+    if started_dt:
+        duration_seconds = round((now - started_dt).total_seconds())
+
+    # Calculate heartbeat age
+    heartbeat_age = None
+    if heartbeat_dt:
+        heartbeat_age = round((now - heartbeat_dt).total_seconds())
+
+    # Resolve state
+    normalized = last_status.lower().strip() if last_status else ""
+    if pid_alive:
+        if heartbeat_age is not None and heartbeat_age > stale_threshold:
+            state = "STALE"
+        else:
+            state = "RUNNING"
+    else:
+        if normalized in ("done", "completed", "complete", "success"):
+            state = "COMPLETED"
+        elif normalized in ("failed", "error", "errored"):
+            state = "FAILED"
+        elif normalized in ("running", "active", "in_progress", "starting"):
+            state = "CRASHED"
+        elif pid is None:
+            state = "UNKNOWN"
+        else:
+            # PID dead, unknown last status
+            state = "CRASHED" if normalized == "" else "UNKNOWN"
+
+    result: dict[str, Any] = {
+        "state": state,
+        "pid_alive": pid_alive,
+    }
+    if started:
+        result["started"] = started
+    if heartbeat:
+        result["last_heartbeat"] = heartbeat
+    if heartbeat_age is not None:
+        result["heartbeat_age_seconds"] = heartbeat_age
+    if duration_seconds is not None:
+        result["duration_seconds"] = duration_seconds
+    return result
+
+
 @app.get("/api/health/processes")
 async def get_process_health(token: Optional[dict] = Depends(auth.get_current_token)):
-    """Get health status of all loki processes (dashboard, session, agents)."""
+    """Get health status of all loki processes (dashboard, session, agents).
+
+    Returns honest state labels: RUNNING, STALE, COMPLETED, FAILED, CRASHED, UNKNOWN.
+    Every entry includes timestamps (started, last_heartbeat, duration_seconds).
+    """
     result: dict[str, Any] = {"dashboard": None, "session": None, "agents": []}
 
     loki_dir = _get_loki_dir()
+    now_iso = datetime.now(timezone.utc).isoformat()
 
     # Dashboard PID
     dpid_file = loki_dir / "dashboard" / "dashboard.pid"
     if dpid_file.exists():
         try:
             dpid = int(dpid_file.read_text().strip())
-            try:
-                os.kill(dpid, 0)
-                result["dashboard"] = {"pid": dpid, "status": "alive"}
-            except OSError:
-                result["dashboard"] = {"pid": dpid, "status": "dead"}
+            state_info = _resolve_process_state(dpid, last_status="running")
+            result["dashboard"] = {"pid": dpid, **state_info}
         except (ValueError, OSError):
             pass
 
@@ -3176,14 +3265,23 @@ async def get_process_health(token: Optional[dict] = Depends(auth.get_current_to
     if spid_file.exists():
         try:
             spid = int(spid_file.read_text().strip())
-            try:
-                os.kill(spid, 0)
-                result["session"] = {"pid": spid, "status": "alive"}
-            except OSError:
-                result["session"] = {"pid": spid, "status": "dead"}
+            state_info = _resolve_process_state(spid, last_status="running")
+            result["session"] = {"pid": spid, **state_info}
         except (ValueError, OSError):
             pass
 
+    # Read dashboard-state.json for heartbeat timestamp
+    state_file = loki_dir / "dashboard-state.json"
+    state_heartbeat = ""
+    if state_file.exists():
+        try:
+            st = os.stat(state_file)
+            state_heartbeat = datetime.fromtimestamp(
+                st.st_mtime, tz=timezone.utc
+            ).isoformat()
+        except OSError:
+            pass
+
     # Agent PIDs
     agents_file = loki_dir / "state" / "agents.json"
     if agents_file.exists():
@@ -3191,18 +3289,21 @@ async def get_process_health(token: Optional[dict] = Depends(auth.get_current_to
             agents = json.loads(agents_file.read_text())
             for agent in agents:
                 pid = agent.get("pid")
-                status = "unknown"
-                if pid:
-                    try:
-                        os.kill(int(pid), 0)
-                        status = "alive"
-                    except (OSError, ValueError):
-                        status = "dead"
+                pid_int = int(pid) if pid else None
+                agent_status = agent.get("status", "")
+                agent_started = agent.get("started", "")
+                agent_heartbeat = agent.get("heartbeat", state_heartbeat)
+                state_info = _resolve_process_state(
+                    pid_int,
+                    last_status=agent_status,
+                    started=agent_started,
+                    heartbeat=agent_heartbeat,
+                )
                 result["agents"].append({
                     "id": agent.get("id", ""),
                     "name": agent.get("name", ""),
                     "pid": pid,
-                    "status": status,
+                    **state_info,
                 })
         except Exception:
             pass
@@ -3216,17 +3317,29 @@ async def get_process_health(token: Optional[dict] = Depends(auth.get_current_to
                 pid_str = entry_file.stem
                 pid = int(pid_str)
                 entry = json.loads(entry_file.read_text())
-                try:
-                    os.kill(pid, 0)
-                    status = "alive"
-                except OSError:
-                    status = "dead"
+                entry_started = entry.get("started", "")
+                entry_heartbeat = entry.get("heartbeat", "")
+                # Use file mtime as heartbeat fallback
+                if not entry_heartbeat:
+                    try:
+                        st = os.stat(entry_file)
+                        entry_heartbeat = datetime.fromtimestamp(
+                            st.st_mtime, tz=timezone.utc
+                        ).isoformat()
+                    except OSError:
+                        pass
+                entry_status = entry.get("status", "running")
+                state_info = _resolve_process_state(
+                    pid,
+                    last_status=entry_status,
+                    started=entry_started,
+                    heartbeat=entry_heartbeat,
+                )
                 registered.append({
                     "pid": pid,
                     "label": entry.get("label", "unknown"),
-                    "started": entry.get("started", ""),
                     "ppid": entry.get("ppid"),
-                    "status": status,
+                    **state_info,
                 })
             except (ValueError, json.JSONDecodeError, OSError):
                 continue
@@ -3234,6 +3347,7 @@ async def get_process_health(token: Optional[dict] = Depends(auth.get_current_to
 
     watchdog_enabled = os.environ.get("LOKI_WATCHDOG", "false").lower() == "true"
     result["watchdog_enabled"] = watchdog_enabled
+    result["checked_at"] = now_iso
 
     return result
 

package/dashboard/static/index.html

@@ -4774,7 +4774,7 @@ var LokiDashboard=(()=>{var X=Object.defineProperty;var gt=Object.getOwnProperty
         <p>Checklist not initialized</p>
         <p class="hint">The PRD checklist will be created during the first iteration when a PRD is provided.</p>
       </div>
-    `}_attachEventListeners(){let t=this.shadowRoot;t&&(t.querySelectorAll(".category-header[data-category]").forEach(e=>{e.addEventListener("click",()=>this._toggleCategory(e.dataset.category))}),t.querySelectorAll("button[data-waive-id]").forEach(e=>{e.addEventListener("click",a=>{a.stopPropagation(),this._waiveItem(e.dataset.waiveId)})}),t.querySelectorAll("button[data-unwaive-id]").forEach(e=>{e.addEventListener("click",a=>{a.stopPropagation(),this._unwaiveItem(e.dataset.unwaiveId)})}))}_escapeHtml(t){return t?String(t).replace(/&/g,"&amp;").replace(/</g,"&lt;").replace(/>/g,"&gt;").replace(/"/g,"&quot;"):""}};customElements.define("loki-checklist-viewer",G);var ht={not_initialized:{color:"var(--loki-text-muted, #71717a)",label:"Not Started",pulse:!1},starting:{color:"var(--loki-yellow, #ca8a04)",label:"Starting...",pulse:!0},running:{color:"var(--loki-green, #16a34a)",label:"Running",pulse:!0},crashed:{color:"var(--loki-red, #dc2626)",label:"Crashed",pulse:!1},stopped:{color:"var(--loki-text-muted, #a1a1aa)",label:"Stopped",pulse:!1}},J=class extends c{static get observedAttributes(){return["api-url","theme"]}constructor(){super(),this._loading=!1,this._error=null,this._api=null,this._pollInterval=null,this._status=null,this._logs=[],this._lastDataHash=null,this._lastLogsHash=null}connectedCallback(){super.connectedCallback(),this._setupApi(),this._loadData(),this._startPolling()}disconnectedCallback(){super.disconnectedCallback(),this._stopPolling()}attributeChangedCallback(t,e,a){e!==a&&(t==="api-url"&&this._api&&(this._api.baseUrl=a,this._loadData()),t==="theme"&&this._applyTheme())}_setupApi(){let t=this.getAttribute("api-url")||window.location.origin;this._api=u({baseUrl:t})}_startPolling(){this._pollInterval=setInterval(()=>this._loadData(),3e3),this._visibilityHandler=()=>{document.hidden?this._pollInterval&&(clearInterval(this._pollInterval),this._pollInterval=null):this._pollInterval||(this._loadData(),this._pollInterval=setInterval(()=>this._loadData(),3e3))},document.addEventListener("visibilitychange",this._visibilityHandler)}_stopPolling(){this._pollInterval&&(clearInterval(this._pollInterval),this._pollInterval=null),this._visibilityHandler&&(document.removeEventListener("visibilitychange",this._visibilityHandler),this._visibilityHandler=null)}async _loadData(){try{let[t,e]=await Promise.all([this._api.getAppRunnerStatus(),this._api.getAppRunnerLogs()]),a=JSON.stringify({status:t?.status,port:t?.port,restarts:t?.restart_count,url:t?.url}),i=JSON.stringify(e?.lines?.slice(-5)||[]),s=i!==this._lastLogsHash;if(a===this._lastDataHash&&!s)return;this._lastDataHash=a,this._lastLogsHash=i,this._status=t,this._logs=e?.lines||[],this._error=null,this.render(),this._scrollLogsToBottom()}catch(t){this._error||(this._error=`Failed to load app status: ${t.message}`,this.render())}}_scrollLogsToBottom(){let t=this.shadowRoot;if(!t)return;let e=t.querySelector(".log-area");e&&(e.scrollTop=e.scrollHeight)}async _handleRestart(){try{await this._api.restartApp(),this._loadData()}catch(t){this._error=`Restart failed: ${t.message}`,this.render()}}async _handleStop(){try{await this._api.stopApp(),this._loadData()}catch(t){this._error=`Stop failed: ${t.message}`,this.render()}}_formatUptime(t){if(!t)return"--";let e=new Date(t),i=Math.floor((new Date-e)/1e3);if(i<60)return`${i}s`;if(i<3600)return`${Math.floor(i/60)}m ${i%60}s`;let s=Math.floor(i/3600),r=Math.floor(i%3600/60);return`${s}h ${r}m`}_isValidUrl(t){if(!t)return!1;try{let e=new URL(t);return e.protocol==="http:"||e.protocol==="https:"}catch{return!1}}_getStyles(){return`
+    `}_attachEventListeners(){let t=this.shadowRoot;t&&(t.querySelectorAll(".category-header[data-category]").forEach(e=>{e.addEventListener("click",()=>this._toggleCategory(e.dataset.category))}),t.querySelectorAll("button[data-waive-id]").forEach(e=>{e.addEventListener("click",a=>{a.stopPropagation(),this._waiveItem(e.dataset.waiveId)})}),t.querySelectorAll("button[data-unwaive-id]").forEach(e=>{e.addEventListener("click",a=>{a.stopPropagation(),this._unwaiveItem(e.dataset.unwaiveId)})}))}_escapeHtml(t){return t?String(t).replace(/&/g,"&amp;").replace(/</g,"&lt;").replace(/>/g,"&gt;").replace(/"/g,"&quot;"):""}};customElements.define("loki-checklist-viewer",G);var ht={not_initialized:{color:"var(--loki-text-muted, #71717a)",label:"Not Started",pulse:!1},starting:{color:"var(--loki-yellow, #ca8a04)",label:"Starting...",pulse:!0},running:{color:"var(--loki-green, #16a34a)",label:"Running",pulse:!0},stale:{color:"var(--loki-yellow, #ca8a04)",label:"Stale",pulse:!1},completed:{color:"var(--loki-text-muted, #a1a1aa)",label:"Completed",pulse:!1},failed:{color:"var(--loki-red, #dc2626)",label:"Failed",pulse:!1},crashed:{color:"var(--loki-red, #dc2626)",label:"Crashed",pulse:!1},stopped:{color:"var(--loki-text-muted, #a1a1aa)",label:"Stopped",pulse:!1},unknown:{color:"var(--loki-text-muted, #71717a)",label:"Unknown",pulse:!1}},J=class extends c{static get observedAttributes(){return["api-url","theme"]}constructor(){super(),this._loading=!1,this._error=null,this._api=null,this._pollInterval=null,this._status=null,this._logs=[],this._lastDataHash=null,this._lastLogsHash=null}connectedCallback(){super.connectedCallback(),this._setupApi(),this._loadData(),this._startPolling()}disconnectedCallback(){super.disconnectedCallback(),this._stopPolling()}attributeChangedCallback(t,e,a){e!==a&&(t==="api-url"&&this._api&&(this._api.baseUrl=a,this._loadData()),t==="theme"&&this._applyTheme())}_setupApi(){let t=this.getAttribute("api-url")||window.location.origin;this._api=u({baseUrl:t})}_startPolling(){this._pollInterval=setInterval(()=>this._loadData(),3e3),this._visibilityHandler=()=>{document.hidden?this._pollInterval&&(clearInterval(this._pollInterval),this._pollInterval=null):this._pollInterval||(this._loadData(),this._pollInterval=setInterval(()=>this._loadData(),3e3))},document.addEventListener("visibilitychange",this._visibilityHandler)}_stopPolling(){this._pollInterval&&(clearInterval(this._pollInterval),this._pollInterval=null),this._visibilityHandler&&(document.removeEventListener("visibilitychange",this._visibilityHandler),this._visibilityHandler=null)}async _loadData(){try{let[t,e]=await Promise.all([this._api.getAppRunnerStatus(),this._api.getAppRunnerLogs()]),a=JSON.stringify({status:t?.status,port:t?.port,restarts:t?.restart_count,url:t?.url}),i=JSON.stringify(e?.lines?.slice(-5)||[]),s=i!==this._lastLogsHash;if(a===this._lastDataHash&&!s)return;this._lastDataHash=a,this._lastLogsHash=i,this._status=t,this._logs=e?.lines||[],this._error=null,this.render(),this._scrollLogsToBottom()}catch(t){this._error||(this._error=`Failed to load app status: ${t.message}`,this.render())}}_scrollLogsToBottom(){let t=this.shadowRoot;if(!t)return;let e=t.querySelector(".log-area");e&&(e.scrollTop=e.scrollHeight)}async _handleRestart(){try{await this._api.restartApp(),this._loadData()}catch(t){this._error=`Restart failed: ${t.message}`,this.render()}}async _handleStop(){try{await this._api.stopApp(),this._loadData()}catch(t){this._error=`Stop failed: ${t.message}`,this.render()}}_formatUptime(t){if(!t)return"--";let e=new Date(t),i=Math.floor((new Date-e)/1e3);if(i<60)return`${i}s`;if(i<3600)return`${Math.floor(i/60)}m ${i%60}s`;let s=Math.floor(i/3600),r=Math.floor(i%3600/60);return`${s}h ${r}m`}_isValidUrl(t){if(!t)return!1;try{let e=new URL(t);return e.protocol==="http:"||e.protocol==="https:"}catch{return!1}}_getStyles(){return`
       .app-status {
         padding: 16px;
         font-family: var(--loki-font-family, system-ui, -apple-system, sans-serif);

package/docs/COMPARISON.md

@@ -37,7 +37,7 @@
 |---------|--------------|-----------|-----------|------------|----------|-----------------|--------------|--------------|
 | **Code Review** | 3 blind reviewers + devil's advocate | Basic | Basic | BugBot PR | Property-based | Artifacts | Doc/Review | Basic |
 | **Anti-Sycophancy** | Yes (CONSENSAGENT) | No | No | No | No | No | No | No |
-| **Quality Gates** | 7 gates + PBT | Basic | Sandbox | Tests | Spec validation | Artifact checks | Tests | Permissions |
+| **Quality Gates** | 9 gates + PBT | Basic | Sandbox | Tests | Spec validation | Artifact checks | Tests | Permissions |
 | **Constitutional AI** | Yes (principles) | No | Refusal training | No | No | No | No | No |
 
 ---
@@ -146,7 +146,7 @@
 
 | Feature | **Zencoder** | **Loki Mode** | **Assessment** |
 |---------|-------------|---------------|----------------|
-| **Four Pillars** | Structured Workflows, SDD, Multi-Agent Verification, Parallel Execution | SDLC + RARV + 7 Gates + Worktrees | TIE |
+| **Four Pillars** | Structured Workflows, SDD, Multi-Agent Verification, Parallel Execution | SDLC + RARV + 9 Gates + Worktrees | TIE |
 | **Spec-Driven Dev** | Specs as first-class objects | OpenAPI-first | TIE |
 | **Multi-Agent Verification** | Model diversity (Claude vs OpenAI, 54% improvement) | 3 blind reviewers + devil's advocate | Different approach (N/A for Claude Code - only Claude models) |
 | **Quality Gates** | Built-in verification loops | 7 explicit gates + anti-sycophancy | **Loki Mode** |
@@ -207,7 +207,7 @@
 | **Skills** | Progressive disclosure | 6 slash commands | N/A | 129 skills | N/A | 35 skills | Memory focus |
 | **Multi-Provider** | Yes (Claude/Codex/Gemini) | 3 CLIs (separate) | No | No | No | No | No |
 | **Memory System** | 3-tier (episodic/semantic/procedural) | None | N/A | N/A | Hybrid | N/A | SQLite+FTS5 |
-| **Quality Gates** | 7 gates + Completion Council | User verify only | Two-Stage Review | N/A | Consensus | Tiered | N/A |
+| **Quality Gates** | 9 gates + Completion Council | User verify only | Two-Stage Review | N/A | Consensus | Tiered | N/A |
 | **Context Mgmt** | Standard | Fresh per task (core innovation) | Fresh per task | N/A | N/A | N/A | Progressive |
 | **Autonomy** | High (minimal human) | Semi (checkpoints) | Human-guided | Human-guided | Orchestrated | Human-guided | N/A |
 
@@ -232,7 +232,7 @@ These are patterns from competing projects that are **practically and scientific
 |----------|---------|-------------------------|
 | **Multi-Provider Support** | Only skill supporting Claude, Codex, and Gemini with graceful degradation | All 8 competitors are Claude-only |
 | **RARV Cycle** | Reason-Act-Reflect-Verify is more rigorous than Plan-Execute | Most use simple Plan-Execute |
-| **7-Gate Quality System** | Static analysis + 3 reviewers + devil's advocate + anti-sycophancy + severity blocking + coverage + debate | Superpowers has 2-stage, others have less |
+| **9-Gate Quality System** | Static analysis + 3 reviewers + devil's advocate + anti-sycophancy + severity blocking + coverage + debate | Superpowers has 2-stage, others have less |
 | **Constitutional AI Integration** | Principles-based self-critique from Anthropic research | None have this |
 | **Anti-Sycophancy (CONSENSAGENT)** | Blind review + devil's advocate prevents groupthink | None have this |
 | **Provider Abstraction Layer** | Clean degradation from full-featured to sequential-only | Claude-only projects can't degrade |
@@ -359,7 +359,7 @@ Tiered agent architecture with explicit escalation:
 |-----------|-------------------|
 | **Autonomy** | Designed for high autonomy with minimal human intervention |
 | **Multi-Agent** | 41 specialized agents in 8 swarms vs 1-8 in competitors |
-| **Quality** | 7 gates + blind review + devil's advocate + property-based testing |
+| **Quality** | 9 gates + blind review + devil's advocate + property-based testing |
 | **Research** | 10+ academic papers integrated vs proprietary/undisclosed |
 | **Anti-Sycophancy** | Only agent with CONSENSAGENT-based blind review |
 | **Memory** | 3-tier memory (episodic/semantic/procedural) + review learning + cross-project |

package/docs/COMPETITIVE-ANALYSIS.md

@@ -20,7 +20,7 @@ GSD is the closest competitor -- a context engineering system that spawns fresh
 | Adoption | 594 stars, 6K/wk npm | 11,903 stars, 21K/wk npm | GSD (20x) |
 | Simplicity | Complex (5.4K-line run.sh, 12 Python modules) | Simple (markdown agents + slash commands) | GSD |
 | Full autonomy | Walk away, come back to deployed product | Human checkpoints at discuss/verify/milestone | Loki |
-| Quality gates | 7-gate + Completion Council + anti-sycophancy | User verification only | Loki |
+| Quality gates | 9-gate + Completion Council + anti-sycophancy | User verification only | Loki |
 | Memory system | Episodic/semantic/procedural + vector search | None | Loki |
 | Context management | Standard | Fresh subagent contexts per task (core innovation) | GSD |
 | Time to value | Learn architecture, understand CLI flags | `npx get-shit-done-cc` and go | GSD |

package/docs/INSTALLATION.md

@@ -2,7 +2,7 @@
 
 The flagship product of [Autonomi](https://www.autonomi.dev/). Complete installation instructions for all platforms and use cases.
 
-**Version:** v5.49.1
+**Version:** v5.49.2
 
 ---
 

package/docs/alternative-installations.md

@@ -0,0 +1,145 @@
+# Alternative Installation Methods
+
+The primary installation method is git clone (see [README](../README.md#installation)). These alternatives serve specific use cases.
+
+---
+
+## npm (Secondary)
+
+**Status**: Working. Version tracks releases automatically.
+
+```bash
+npm install -g loki-mode
+```
+
+**Limitation**: Installs to `node_modules`, not `~/.claude/skills/`. To use as a Claude Code skill, you must symlink:
+
+```bash
+npm install -g loki-mode
+ln -sf "$(npm root -g)/loki-mode" ~/.claude/skills/loki-mode
+```
+
+**Best for**: CI/CD pipelines, programmatic access via `loki` CLI.
+
+---
+
+## Homebrew (Secondary)
+
+**Status**: Working. Tap and formula exist, version current.
+
+```bash
+brew tap asklokesh/tap
+brew install loki-mode
+```
+
+**Limitation**: Installs the `loki` CLI binary only. Does NOT install the Claude Code skill. To use with Claude Code, also run:
+
+```bash
+git clone https://github.com/asklokesh/loki-mode.git ~/.claude/skills/loki-mode
+```
+
+**Best for**: Users who want the `loki` CLI wrapper for autonomous mode (`loki start`, `loki stop`, `loki cleanup`).
+
+---
+
+## Docker (Secondary)
+
+**Status**: Image exists on Docker Hub. Tags: `latest`, version-specific (e.g., `5.49.1`).
+
+```bash
+docker pull asklokesh/loki-mode:latest
+```
+
+**Limitation**: Claude Code is an interactive CLI that requires API keys and terminal access. Running it inside a Docker container is not the standard workflow. Docker is useful for:
+
+- CI/CD sandbox execution (running `loki` in isolated environments)
+- Testing Loki Mode without modifying your local system
+- Air-gapped environments with pre-built images
+
+**Not recommended for**: Interactive Claude Code sessions. Use the git clone method instead.
+
+See [DOCKER_README.md](../DOCKER_README.md) for Docker-specific usage instructions.
+
+---
+
+## GitHub Action (Secondary)
+
+**Status**: Working. Adds automated AI code review to pull requests.
+
+```yaml
+# .github/workflows/loki-review.yml
+name: Loki Code Review
+on:
+  pull_request:
+    types: [opened, synchronize]
+permissions:
+  contents: read
+  pull-requests: write
+jobs:
+  review:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: asklokesh/loki-mode@v5
+        with:
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+          mode: review
+          provider: claude
+          max_iterations: 3
+          budget_limit: '5.00'
+        env:
+          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+```
+
+**Prerequisites:**
+- API key for your provider (set as repository secret): `ANTHROPIC_API_KEY`, `OPENAI_API_KEY`, or `GOOGLE_API_KEY`
+- The action auto-installs `loki-mode` and `@anthropic-ai/claude-code`
+
+**Action Inputs:**
+
+| Input | Default | Description |
+|-------|---------|-------------|
+| `mode` | `review` | `review`, `fix`, or `test` |
+| `provider` | `claude` | `claude`, `codex`, or `gemini` |
+| `budget_limit` | `5.00` | Max cost in USD |
+| `max_iterations` | `3` | Max RARV cycles |
+| `github_token` | (required) | GitHub token for PR comments |
+| `prd_file` | | Path to PRD file (for fix/test modes) |
+
+**Modes:**
+
+| Mode | Description |
+|------|-------------|
+| `review` | Analyze PR diff, post structured review as PR comment |
+| `fix` | Automatically fix issues found in the codebase |
+| `test` | Run autonomous test generation and validation |
+
+**Best for**: Automated PR review and CI/CD integration.
+
+---
+
+## GitHub Release Download (Secondary)
+
+**Status**: Working. Release assets available for each version.
+
+```bash
+# Download and extract to skills directory
+curl -sL https://github.com/asklokesh/loki-mode/archive/refs/tags/v5.49.1.tar.gz | tar xz
+mv loki-mode-5.49.1 ~/.claude/skills/loki-mode
+```
+
+**Best for**: Offline or air-gapped environments, pinned version deployments.
+
+---
+
+## VS Code Extension (Secondary)
+
+**Status**: Available on VS Code Marketplace.
+
+Search for "Loki Mode" in VS Code Extensions, or:
+
+```bash
+code --install-extension asklokesh.loki-mode
+```
+
+**Best for**: VS Code users who want dashboard integration within their editor.

package/docs/cursor-comparison.md

@@ -11,7 +11,7 @@
 |-----------|--------|-----------|--------|
 | **Proven Scale** | 1M+ LoC, large agent count | Benchmarks only | Cursor |
 | **Research Foundation** | Empirical iteration | 25+ academic citations | Loki Mode |
-| **Quality Assurance** | Workers self-manage | 7-gate system + anti-sycophancy | Loki Mode |
+| **Quality Assurance** | Workers self-manage | 9-gate system + anti-sycophancy | Loki Mode |
 | **Anti-Sycophancy** | Not mentioned | CONSENSAGENT blind review | Loki Mode |
 | **Velocity-Quality Balance** | Not mentioned | arXiv-backed metrics | Loki Mode |
 | **Full SDLC Coverage** | Code generation focus | PRD to production + growth | Loki Mode |
@@ -66,7 +66,7 @@ velocity_quality_balance:
 
 ---
 
-### 3. 7-Gate Quality System
+### 3. 9-Gate Quality System
 
 **Loki Mode's Gates:**
 1. Input Guardrails - Validate scope, detect injection (OpenAI SDK pattern)
@@ -174,7 +174,7 @@ Cursor learned through failure:
 ### 3. Simplicity Principle
 > "A surprising amount of the system's behavior comes down to how we prompt the agents. The harness and models matter, but the prompts matter more."
 
-**Loki Mode:** More complex infrastructure (7 gates, 41 agent types, memory systems). May be over-engineered for some use cases.
+**Loki Mode:** More complex infrastructure (9 gates, 41 agent types, memory systems). May be over-engineered for some use cases.
 
 ---
 
@@ -192,7 +192,7 @@ We incorporated Cursor's proven patterns:
 ## Conclusion
 
 **Loki Mode is scientifically better in:**
-- Quality assurance (research-backed 7-gate system)
+- Quality assurance (research-backed 9-gate system)
 - Anti-sycophancy (CONSENSAGENT blind review)
 - Velocity-quality balance (arXiv metrics)
 - Full SDLC coverage (PRD to growth)

package/mcp/__init__.py

@@ -21,4 +21,4 @@ try:
 except ImportError:
     __all__ = ['mcp']
 
-__version__ = '5.49.1'
+__version__ = '5.49.2'

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "loki-mode",
-  "version": "5.49.1",
+  "version": "5.49.2",
   "description": "Loki Mode by Autonomi - Multi-agent autonomous startup system for Claude Code, Codex CLI, and Gemini CLI",
   "keywords": [
     "autonomi",

package/references/core-workflow.md

@@ -6,7 +6,7 @@ Full RARV cycle, CONTINUITY.md template, and autonomy rules.
 
 ## Autonomy Rules
 
-**This system runs with ZERO human intervention.**
+**This system runs with minimal human intervention.** Human oversight is expected for deployment credentials, domain setup, API keys, and critical business decisions.
 
 ### Core Rules
 1. **NEVER ask questions** - Do not say "Would you like me to...", "Should I...", or "What would you prefer?"

package/skills/00-index.md

@@ -41,7 +41,7 @@
 
 ### quality-gates.md
 **When:** Code review, pre-commit checks, quality assurance
-- 7-gate quality system
+- 9-gate quality system
 - Blind review + anti-sycophancy
 - Velocity-quality feedback loop (arXiv research)
 - Mandatory quality checks per task

package/skills/artifacts.md

@@ -36,7 +36,7 @@ format: "markdown"
 contents:
   - Phase name and duration
   - Tasks completed (from queue)
-  - Quality gate results (7 gates)
+  - Quality gate results (9 gates)
   - Coverage metrics
   - Known issues / TODOs
 ```

package/skills/quality-gates.md

@@ -2,7 +2,7 @@
 
 **Never ship code without passing all quality gates.**
 
-## The 7 Quality Gates
+## The 9 Quality Gates
 
 1. **Input Guardrails** - Validate scope, detect injection, check constraints (OpenAI SDK)
 2. **Static Analysis** - CodeQL, ESLint/Pylint, type checking
@@ -11,6 +11,8 @@
 5. **Output Guardrails** - Validate code quality, spec compliance, no secrets (tripwire on fail)
 6. **Severity-Based Blocking** - Critical/High/Medium = BLOCK; Low/Cosmetic = TODO comment
 7. **Test Coverage Gates** - Unit: 100% pass, >80% coverage; Integration: 100% pass
+8. **Mock Detector** - Classifies internal vs external mocks; flags tests that never import source code, tautological assertions, and high internal mock ratios
+9. **Test Mutation Detector** - Detects assertion value changes alongside implementation changes (test fitting), low assertion density, and missing pass/fail tracking
 
 ## Guardrails Execution Modes