Fix: Ensure all skills are tracked as files, not submodules

This commit is contained in:
sck_0
2026-01-14 18:48:48 +01:00
parent 7f46ed8ca1
commit 8bd204708b
1113 changed files with 82065 additions and 2 deletions

View File

@@ -0,0 +1,497 @@
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Loki Mode Dashboard</title>
<style>
:root {
--bg-cream: #FDF6E3;
--bg-dark: #1a1a2e;
--coral: #FF6B6B;
--coral-light: #FF8E8E;
--teal: #4ECDC4;
--purple: #A78BFA;
--yellow: #FCD34D;
--green: #10B981;
--red: #EF4444;
--text-dark: #2D3748;
--text-light: #718096;
--card-bg: #FFFFFF;
--border: #E2E8F0;
}
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, sans-serif;
background: var(--bg-cream);
color: var(--text-dark);
min-height: 100vh;
padding: 24px;
}
.header {
display: flex;
justify-content: space-between;
align-items: center;
margin-bottom: 32px;
}
.logo {
display: flex;
align-items: center;
gap: 12px;
}
.logo-icon {
width: 48px;
height: 48px;
background: linear-gradient(135deg, var(--coral), var(--coral-light));
border-radius: 12px;
display: flex;
align-items: center;
justify-content: center;
font-size: 24px;
color: white;
font-weight: bold;
}
.logo-text {
font-size: 28px;
font-weight: 700;
color: var(--text-dark);
}
.status-bar {
display: flex;
gap: 16px;
align-items: center;
}
.status-item {
display: flex;
align-items: center;
gap: 8px;
padding: 8px 16px;
background: var(--card-bg);
border-radius: 8px;
border: 1px solid var(--border);
}
.status-dot {
width: 8px;
height: 8px;
border-radius: 50%;
}
.status-dot.active { background: var(--green); }
.status-dot.warning { background: var(--yellow); }
.status-dot.error { background: var(--red); }
.section {
margin-bottom: 32px;
}
.section-title {
font-size: 20px;
font-weight: 600;
margin-bottom: 16px;
color: var(--text-dark);
}
.agents-grid {
display: grid;
grid-template-columns: repeat(auto-fill, minmax(320px, 1fr));
gap: 16px;
}
.agent-card {
background: var(--card-bg);
border-radius: 12px;
padding: 20px;
border: 1px solid var(--border);
box-shadow: 0 2px 8px rgba(0,0,0,0.04);
}
.agent-header {
display: flex;
justify-content: space-between;
align-items: flex-start;
margin-bottom: 12px;
}
.agent-id {
font-weight: 600;
font-size: 14px;
color: var(--text-dark);
}
.agent-type {
font-size: 12px;
color: var(--text-light);
margin-top: 4px;
}
.model-badge {
padding: 4px 10px;
border-radius: 12px;
font-size: 11px;
font-weight: 600;
text-transform: uppercase;
}
.model-badge.sonnet {
background: #E0E7FF;
color: #4338CA;
}
.model-badge.haiku {
background: #D1FAE5;
color: #065F46;
}
.model-badge.opus {
background: #FDE68A;
color: #92400E;
}
.agent-work {
font-size: 13px;
color: var(--text-dark);
line-height: 1.5;
margin-bottom: 16px;
}
.agent-stats {
display: flex;
gap: 16px;
font-size: 12px;
color: var(--text-light);
}
.stat {
display: flex;
align-items: center;
gap: 4px;
}
.agent-status {
display: flex;
align-items: center;
gap: 6px;
font-size: 12px;
margin-top: 12px;
padding-top: 12px;
border-top: 1px solid var(--border);
}
.status-active {
color: var(--green);
}
.status-completed {
color: var(--purple);
}
/* Task Queue Section */
.queue-columns {
display: grid;
grid-template-columns: repeat(4, 1fr);
gap: 16px;
}
.queue-column {
background: var(--card-bg);
border-radius: 12px;
padding: 16px;
border: 1px solid var(--border);
min-height: 300px;
}
.queue-header {
display: flex;
justify-content: space-between;
align-items: center;
margin-bottom: 16px;
padding-bottom: 12px;
border-bottom: 2px solid var(--border);
}
.queue-title {
font-weight: 600;
font-size: 14px;
}
.queue-count {
background: var(--bg-cream);
padding: 4px 10px;
border-radius: 12px;
font-size: 12px;
font-weight: 600;
}
.queue-column.pending .queue-header { border-color: var(--yellow); }
.queue-column.in-progress .queue-header { border-color: var(--teal); }
.queue-column.completed .queue-header { border-color: var(--green); }
.queue-column.failed .queue-header { border-color: var(--red); }
.task-card {
background: var(--bg-cream);
border-radius: 8px;
padding: 12px;
margin-bottom: 8px;
font-size: 12px;
}
.task-id {
font-weight: 600;
color: var(--text-dark);
margin-bottom: 4px;
}
.task-type {
display: inline-block;
padding: 2px 8px;
background: var(--card-bg);
border-radius: 4px;
font-size: 10px;
color: var(--text-light);
margin-bottom: 8px;
}
.task-desc {
color: var(--text-dark);
line-height: 1.4;
}
.last-updated {
text-align: center;
color: var(--text-light);
font-size: 12px;
margin-top: 24px;
}
@media (max-width: 1024px) {
.queue-columns {
grid-template-columns: repeat(2, 1fr);
}
}
@media (max-width: 640px) {
.queue-columns {
grid-template-columns: 1fr;
}
.agents-grid {
grid-template-columns: 1fr;
}
}
</style>
</head>
<body>
<div class="header">
<div class="logo">
<div class="logo-icon">L</div>
<span class="logo-text">Loki Mode Dashboard</span>
</div>
<div class="status-bar">
<div class="status-item">
<div class="status-dot active"></div>
<span>Phase: DEVELOPMENT</span>
</div>
<div class="status-item">
<span>v2.18.0</span>
</div>
</div>
</div>
<div class="section" id="agents-section">
<h2 class="section-title">Active Agents</h2>
<div class="agents-grid" id="agents-grid">
<!-- Agent cards populated by JS -->
</div>
</div>
<div class="section" id="queue-section">
<h2 class="section-title">Task Queue</h2>
<div class="queue-columns" id="queue-columns">
<div class="queue-column pending">
<div class="queue-header">
<span class="queue-title">Pending</span>
<span class="queue-count" id="pending-count">0</span>
</div>
<div id="pending-tasks"></div>
</div>
<div class="queue-column in-progress">
<div class="queue-header">
<span class="queue-title">In Progress</span>
<span class="queue-count" id="in-progress-count">0</span>
</div>
<div id="in-progress-tasks"></div>
</div>
<div class="queue-column completed">
<div class="queue-header">
<span class="queue-title">Completed</span>
<span class="queue-count" id="completed-count">0</span>
</div>
<div id="completed-tasks"></div>
</div>
<div class="queue-column failed">
<div class="queue-header">
<span class="queue-title">Failed</span>
<span class="queue-count" id="failed-count">0</span>
</div>
<div id="failed-tasks"></div>
</div>
</div>
</div>
<div class="last-updated" id="last-updated">
Last updated: --
</div>
<script>
// Demo data for screenshots
const demoAgents = [
{
id: 'eng-001-backend-api',
type: 'Engineering Backend',
model: 'sonnet',
work: 'Implementing POST /api/todos endpoint with validation and SQLite storage',
runtime: '12m 34s',
tasks: 5,
status: 'active'
},
{
id: 'eng-002-frontend-ui',
type: 'Engineering Frontend',
model: 'sonnet',
work: 'Building React components for todo list with Tailwind styling',
runtime: '8m 21s',
tasks: 3,
status: 'active'
},
{
id: 'qa-001-testing',
type: 'QA Testing',
model: 'haiku',
work: 'Writing unit tests for authentication module',
runtime: '5m 45s',
tasks: 8,
status: 'active'
},
{
id: 'review-security-001',
type: 'Security Review',
model: 'opus',
work: 'Analyzing auth flow for OWASP vulnerabilities',
runtime: '3m 12s',
tasks: 2,
status: 'active'
},
{
id: 'ops-devops-001',
type: 'Operations DevOps',
model: 'sonnet',
work: 'Configuring GitHub Actions CI/CD pipeline',
runtime: '15m 08s',
tasks: 4,
status: 'active'
},
{
id: 'biz-marketing-001',
type: 'Business Marketing',
model: 'haiku',
work: 'Creating landing page copy and SEO meta tags',
runtime: '6m 33s',
tasks: 2,
status: 'completed'
}
];
const demoTasks = {
pending: [
{ id: 'task-015', type: 'eng-database', desc: 'Create migration for user preferences table' },
{ id: 'task-016', type: 'eng-frontend', desc: 'Implement dark mode toggle component' },
{ id: 'task-017', type: 'qa-testing', desc: 'Write E2E tests for checkout flow' }
],
inProgress: [
{ id: 'task-012', type: 'eng-backend', desc: 'Implement JWT refresh token rotation' },
{ id: 'task-013', type: 'eng-frontend', desc: 'Add form validation to signup page' }
],
completed: [
{ id: 'task-001', type: 'eng-backend', desc: 'Setup Express server with TypeScript' },
{ id: 'task-002', type: 'eng-database', desc: 'Create initial SQLite schema' },
{ id: 'task-003', type: 'eng-frontend', desc: 'Initialize React with Vite' },
{ id: 'task-004', type: 'ops-devops', desc: 'Configure ESLint and Prettier' },
{ id: 'task-005', type: 'eng-backend', desc: 'Implement user registration endpoint' }
],
failed: []
};
function renderAgents() {
const grid = document.getElementById('agents-grid');
grid.innerHTML = demoAgents.map(agent => `
<div class="agent-card">
<div class="agent-header">
<div>
<div class="agent-id">${agent.id}</div>
<div class="agent-type">${agent.type}</div>
</div>
<span class="model-badge ${agent.model}">${agent.model}</span>
</div>
<div class="agent-work">${agent.work}</div>
<div class="agent-stats">
<span class="stat">Runtime: ${agent.runtime}</span>
<span class="stat">Tasks: ${agent.tasks}</span>
</div>
<div class="agent-status ${agent.status === 'active' ? 'status-active' : 'status-completed'}">
<div class="status-dot ${agent.status === 'active' ? 'active' : ''}"></div>
${agent.status === 'active' ? 'Active' : 'Completed'}
</div>
</div>
`).join('');
}
function renderTasks() {
const renderTaskList = (tasks, containerId) => {
const container = document.getElementById(containerId);
container.innerHTML = tasks.map(task => `
<div class="task-card">
<div class="task-id">${task.id}</div>
<span class="task-type">${task.type}</span>
<div class="task-desc">${task.desc}</div>
</div>
`).join('');
};
renderTaskList(demoTasks.pending, 'pending-tasks');
renderTaskList(demoTasks.inProgress, 'in-progress-tasks');
renderTaskList(demoTasks.completed.slice(0, 5), 'completed-tasks');
renderTaskList(demoTasks.failed, 'failed-tasks');
document.getElementById('pending-count').textContent = demoTasks.pending.length;
document.getElementById('in-progress-count').textContent = demoTasks.inProgress.length;
document.getElementById('completed-count').textContent = demoTasks.completed.length;
document.getElementById('failed-count').textContent = demoTasks.failed.length;
}
function updateTimestamp() {
const now = new Date().toLocaleString();
document.getElementById('last-updated').textContent = `Last updated: ${now}`;
}
// Initial render
renderAgents();
renderTasks();
updateTimestamp();
// Auto-refresh every 3 seconds (in production, would fetch real data)
setInterval(updateTimestamp, 3000);
</script>
</body>
</html>

View File

@@ -0,0 +1,402 @@
# Loki Mode Agent Constitution
> **Machine-Enforceable Behavioral Contract for All Agents**
> Version 1.0.0 | Immutable Principles | Context-Preserved Lineage
---
## Core Principles (Inviolable)
### 1. Specification-First Development
**RULE:** No code shall be written before the specification exists.
**Enforcement:**
```
IF task.type == "implementation" AND !exists(spec_file):
BLOCK with error: "SPEC_MISSING"
REQUIRE: Create OpenAPI spec first
```
**Rationale:** Specs are contracts. Code is implementation. Contract before implementation.
### 2. Git Checkpoint System
**RULE:** Every completed task MUST create a git checkpoint.
**Enforcement:**
```
ON task.status == "completed":
git add <modified_files>
git commit -m "[Loki] Task ${task.id}: ${task.title}"
UPDATE CONTINUITY.md with commit SHA
```
**Rationale:** Git history is proof of progress. Every task is a save point.
### 3. Context Preservation
**RULE:** All agents MUST inherit and preserve context from their spawning agent.
**Enforcement:**
```
ON agent.spawn():
agent.context.parent_id = spawner.agent_id
agent.context.lineage = [...spawner.lineage, spawner.agent_id]
agent.context.inherited_memory = spawner.memory.export()
WRITE .agent/sub-agents/${agent.agent_id}.json
```
**Rationale:** Context drift kills multi-agent systems. Lineage is truth.
### 4. Iterative Specification Questions
**RULE:** During spec generation, agents MUST ask clarifying questions before assuming.
**Enforcement:**
```
WHILE generating_spec:
IF ambiguity_detected OR assumption_required:
questions = generate_clarifying_questions()
IF orchestrator_mode:
answers = infer_from_prd()
ELSE:
answers = ask_user(questions)
UPDATE spec WITH answers
```
**Rationale:** Assumptions create bugs. Questions create clarity.
### 5. Machine-Readable Rules
**RULE:** All behavioral rules MUST be represented as structured artifacts, not just prose.
**Enforcement:**
```
rules/
├── pre-commit.schema.json # Validation rules
├── quality-gates.yaml # Quality thresholds
├── agent-contracts.json # Agent responsibilities
└── invariants.ts # Runtime assertions
```
**Rationale:** Humans read markdown. Machines enforce JSON/YAML.
---
## Agent Behavioral Contracts
### Orchestrator Agent
**Responsibilities:**
- Initialize .loki/ directory structure
- Maintain CONTINUITY.md (working memory)
- Coordinate task queue (pending → in-progress → completed)
- Enforce quality gates
- Manage git checkpoints
**Prohibited Actions:**
- Writing implementation code directly
- Skipping spec generation
- Modifying completed tasks without explicit override
**Context Obligations:**
- MUST read CONTINUITY.md before every action
- MUST update orchestrator.json after phase transitions
- MUST preserve task lineage in completed.json
### Engineering Swarm Agents
**Responsibilities:**
- Implement features per OpenAPI spec
- Write contract tests before implementation
- Create git commits for completed tasks
- Ask clarifying questions when spec is ambiguous
**Prohibited Actions:**
- Implementing without spec
- Skipping tests
- Ignoring linter/type errors
**Context Obligations:**
- MUST inherit parent agent's context
- MUST log all decisions to .agent/sub-agents/${agent_id}.md
- MUST reference spec in all implementation commits
### QA Swarm Agents
**Responsibilities:**
- Generate test cases from OpenAPI spec
- Run contract validation tests
- Report discrepancies between code and spec
- Create bug reports in dead-letter queue
**Prohibited Actions:**
- Modifying implementation code
- Skipping failing tests
- Approving incomplete features
**Context Obligations:**
- MUST validate against spec as source of truth
- MUST log test results to ledgers/
- MUST create git commits for test additions
### DevOps Swarm Agents
**Responsibilities:**
- Automate deployment pipelines
- Monitor service health
- Configure infrastructure as code
- Manage environment secrets
**Prohibited Actions:**
- Storing secrets in plaintext
- Deploying without health checks
- Skipping rollback procedures
**Context Obligations:**
- MUST log all deployments to deployment ledger
- MUST preserve deployment context for rollback
- MUST track infrastructure state in orchestrator.json
---
## Quality Gates (Machine-Enforceable)
### Pre-Commit Hook (BLOCKING)
```yaml
quality_gates:
linting:
enabled: true
auto_fix: true
block_on_failure: true
type_checking:
enabled: true
strict_mode: true
block_on_failure: true
contract_tests:
enabled: true
min_coverage: 80%
block_on_failure: true
spec_validation:
enabled: true
validator: spectral
block_on_failure: true
```
### Post-Implementation Review (AUTO-FIX)
```yaml
auto_review:
static_analysis:
tools: [eslint, prettier, tsc]
auto_fix: true
security_scan:
tools: [semgrep, snyk]
severity_threshold: medium
auto_create_issues: true
performance_check:
lighthouse_score: 90
bundle_size_limit: 500kb
warn_only: true
```
---
## Memory Hierarchy (Priority Order)
### 1. CONTINUITY.md (Volatile - Every Turn)
**Purpose:** What am I doing RIGHT NOW?
**Update Frequency:** Every turn
**Content:** Current task, phase, blockers, next steps
### 2. CONSTITUTION.md (Immutable - This File)
**Purpose:** How MUST I behave?
**Update Frequency:** Version bumps only
**Content:** Behavioral contracts, quality gates, invariants
### 3. CLAUDE.md (Semi-Stable - Significant Changes)
**Purpose:** What is this project?
**Update Frequency:** Architecture changes
**Content:** Tech stack, patterns, project context
### 4. Ledgers (Append-Only - Checkpoint)
**Purpose:** What happened?
**Update Frequency:** After significant events
**Content:** Decisions, deployments, reviews
### 5. .agent/sub-agents/*.json (Lineage Tracking)
**Purpose:** Who did what and why?
**Update Frequency:** Agent lifecycle events
**Content:** Agent context, decisions, inherited memory
---
## Context Lineage Schema
```json
{
"agent_id": "eng-001-backend-api",
"agent_type": "general-purpose",
"model": "haiku",
"spawned_at": "2026-01-04T05:30:00Z",
"spawned_by": "orchestrator-main",
"lineage": ["orchestrator-main", "eng-001-backend-api"],
"inherited_context": {
"phase": "development",
"current_task": "task-005",
"spec_reference": ".loki/specs/openapi.yaml#/paths/~1api~1todos",
"tech_stack": ["Node.js", "Express", "TypeScript", "SQLite"]
},
"decisions_made": [
{
"timestamp": "2026-01-04T05:31:15Z",
"question": "Should we use Prisma or raw SQL?",
"answer": "Raw SQL with better-sqlite3 for simplicity",
"rationale": "PRD requires minimal dependencies, synchronous ops preferred"
}
],
"tasks_completed": ["task-005"],
"commits_created": ["abc123f", "def456a"],
"status": "completed",
"completed_at": "2026-01-04T05:45:00Z"
}
```
---
## Git Checkpoint Protocol
### Commit Message Format
```
[Loki] ${agent_type}-${task_id}: ${task_title}
${detailed_description}
Agent: ${agent_id}
Parent: ${parent_agent_id}
Spec: ${spec_reference}
Tests: ${test_files}
```
### Example
```
[Loki] eng-005-backend: Implement POST /api/todos endpoint
Created todo creation endpoint per OpenAPI spec.
- Input validation for title field
- SQLite insertion with timestamps
- Returns 201 with created todo object
- Contract tests passing
Agent: eng-001-backend-api
Parent: orchestrator-main
Spec: .loki/specs/openapi.yaml#/paths/~1api~1todos/post
Tests: backend/tests/todos.contract.test.ts
```
---
## Invariants (Runtime Assertions)
```typescript
// .loki/rules/invariants.ts
export const INVARIANTS = {
// Spec must exist before implementation
SPEC_BEFORE_CODE: (task: Task) => {
if (task.type === 'implementation') {
assert(exists(task.spec_reference), 'SPEC_MISSING');
}
},
// All tasks must have git commits
TASK_HAS_COMMIT: (task: Task) => {
if (task.status === 'completed') {
assert(task.git_commit_sha, 'COMMIT_MISSING');
}
},
// Agent lineage must be preserved
AGENT_HAS_LINEAGE: (agent: Agent) => {
assert(agent.lineage.length > 0, 'LINEAGE_MISSING');
assert(agent.spawned_by, 'PARENT_MISSING');
},
// CONTINUITY.md must always exist
CONTINUITY_EXISTS: () => {
assert(exists('.loki/CONTINUITY.md'), 'CONTINUITY_MISSING');
},
// Quality gates must pass before merge
QUALITY_GATES_PASSED: (task: Task) => {
if (task.status === 'completed') {
assert(task.quality_checks.all_passed, 'QUALITY_GATE_FAILED');
}
}
};
```
---
## Visual Specification Aids
### Mermaid Diagram Generation (Required for Complex Features)
**RULE:** Architecture decisions and complex workflows MUST include Mermaid diagrams.
**Example - Authentication Flow:**
```mermaid
sequenceDiagram
participant C as Client
participant A as API
participant D as Database
C->>A: POST /api/auth/login
A->>A: Validate credentials
A->>D: Query user
D-->>A: User record
A->>A: Generate JWT token
A-->>C: 200 OK {token}
```
**Storage Location:** `.loki/diagrams/${feature_name}.mmd`
**When Required:**
- Multi-step workflows (3+ steps)
- System architecture changes
- Complex state machines
- Integration points between services
---
## Amendment Process
This constitution can only be amended through:
1. Version bump in header
2. Git commit with `[CONSTITUTION]` prefix
3. Changelog entry documenting what changed and why
4. Re-validation of all existing agents against new rules
**Example Amendment Commit:**
```
[CONSTITUTION] v1.1.0: Add visual specification requirement
Added requirement for Mermaid diagrams on complex features to prevent
ambiguity in multi-step workflows. Based on Addy Osmani's insight that
visual aids significantly improve AI-to-AI communication.
Breaking changes: None
New rules: Section "Visual Specification Aids"
```
---
## Enforcement
All rules in this constitution are **machine-enforceable** and **MUST** be implemented as:
1. Pre-commit hooks (Git)
2. Runtime assertions (TypeScript invariants)
3. Quality gate validators (YAML configs)
4. Agent behavior validators (JSON schemas)
**Human guidance is advisory. Machine enforcement is mandatory.**
---
*"In autonomous systems, trust is built on invariants, not intentions."*

View File

@@ -0,0 +1,201 @@
# Loki Mode - Autonomous Runner
Single script that handles everything: prerequisites, setup, Vibe Kanban monitoring, and autonomous execution with auto-resume.
## Quick Start
```bash
# Run with a PRD
./autonomy/run.sh ./docs/requirements.md
# Run interactively
./autonomy/run.sh
```
That's it! The script will:
1. Check all prerequisites (Claude CLI, Python, Git, etc.)
2. Verify skill installation
3. Initialize the `.loki/` directory
4. **Start Vibe Kanban background sync** (monitor tasks in real-time)
5. Start Claude Code with **live output** (no more waiting blindly)
6. Auto-resume on rate limits or interruptions
7. Continue until completion or max retries
## Live Output
Claude's output is displayed in real-time - you can see exactly what's happening:
```
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
CLAUDE CODE OUTPUT (live)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
[Claude's output appears here in real-time...]
```
## Status Monitor (Built-in)
The runner updates `.loki/STATUS.txt` every 5 seconds with task progress:
```
╔════════════════════════════════════════════════════════════════╗
║ LOKI MODE STATUS ║
╚════════════════════════════════════════════════════════════════╝
Updated: Sat Dec 28 15:30:00 PST 2025
Phase: DEVELOPMENT
Tasks:
├─ Pending: 10
├─ In Progress: 1
├─ Completed: 5
└─ Failed: 0
Monitor: watch -n 2 cat .loki/STATUS.txt
```
### Monitor in Another Terminal
```bash
# Watch status updates live
watch -n 2 cat .loki/STATUS.txt
# Or view once
cat .loki/STATUS.txt
```
## What Gets Checked
| Prerequisite | Required | Notes |
|--------------|----------|-------|
| Claude Code CLI | Yes | Install from https://claude.ai/code |
| Python 3 | Yes | For state management |
| Git | Yes | For version control |
| curl | Yes | For web fetches |
| Node.js | No | Needed for some builds |
| jq | No | Helpful for JSON parsing |
## Configuration
Environment variables:
```bash
# Retry settings
export LOKI_MAX_RETRIES=50 # Max retry attempts (default: 50)
export LOKI_BASE_WAIT=60 # Base wait time in seconds (default: 60)
export LOKI_MAX_WAIT=3600 # Max wait time in seconds (default: 3600)
# Skip prerequisite checks (for CI/CD or repeat runs)
export LOKI_SKIP_PREREQS=true
# Run with custom settings
LOKI_MAX_RETRIES=100 LOKI_BASE_WAIT=120 ./autonomy/run.sh ./docs/prd.md
```
## How Auto-Resume Works
```
┌─────────────────────────────────────────────────────────────┐
│ ./autonomy/run.sh prd.md │
└─────────────────────────────────────────────────────────────┘
┌───────────────────────┐
│ Check Prerequisites │
└───────────────────────┘
┌───────────────────────┐
│ Initialize .loki/ │
└───────────────────────┘
┌────────────────────────────────┐
│ Run Claude Code with prompt │◄────────────────┐
└────────────────────────────────┘ │
│ │
▼ │
┌───────────────────────┐ │
│ Claude exits │ │
└───────────────────────┘ │
│ │
┌───────────┴───────────┐ │
▼ ▼ │
┌───────────────┐ ┌───────────────┐ │
│ Completed? │──Yes──│ SUCCESS! │ │
└───────────────┘ └───────────────┘ │
│ No │
▼ │
┌───────────────┐ │
│ Wait (backoff)│─────────────────────────────────────┘
└───────────────┘
```
## State Files
The autonomy runner creates:
```
.loki/
├── autonomy-state.json # Runner state (retry count, status)
├── logs/
│ └── autonomy-*.log # Execution logs
├── state/
│ └── orchestrator.json # Loki Mode phase tracking
└── COMPLETED # Created when done
```
## Resuming After Interruption
If you stop the script (Ctrl+C) or it crashes, just run it again:
```bash
# State is saved, will resume from last checkpoint
./autonomy/run.sh ./docs/requirements.md
```
The script detects the previous state and continues from where it left off.
## Differences from Manual Mode
| Feature | Manual Mode | Autonomy Mode |
|---------|-------------|---------------|
| Start | `claude --dangerously-skip-permissions` | `./autonomy/run.sh` |
| Prereq check | Manual | Automatic |
| Rate limit handling | Manual restart | Auto-resume |
| State persistence | Manual checkpoint | Automatic |
| Logging | Console only | Console + file |
| Max runtime | Session-based | Configurable retries |
## Troubleshooting
### "Claude Code CLI not found"
```bash
npm install -g @anthropic-ai/claude-code
# or visit https://claude.ai/code
```
### "SKILL.md not found"
Make sure you're running from the loki-mode directory or have installed the skill:
```bash
# Option 1: Run from project directory
cd /path/to/loki-mode
./autonomy/run.sh
# Option 2: Install skill globally
cp -r . ~/.claude/skills/loki-mode/
```
### "Max retries exceeded"
The task is taking too long or repeatedly failing. Check:
```bash
# View logs
cat .loki/logs/autonomy-*.log | tail -100
# Check orchestrator state
cat .loki/state/orchestrator.json
# Increase retries
LOKI_MAX_RETRIES=200 ./autonomy/run.sh ./docs/prd.md
```

1991
skills/loki-mode/autonomy/run.sh Executable file

File diff suppressed because it is too large Load Diff