{"id":46448034,"url":"https://github.com/kujirafu/agentrune","last_synced_at":"2026-04-02T14:02:13.230Z","repository":{"id":342047451,"uuid":"1171506853","full_name":"Kujirafu/AgentRune","owner":"Kujirafu","description":"Multi-session AI agent orchestration — desktop + mobile command center for Claude Code, Codex \u0026 more. Skill chains, shared memory, smart dispatch.","archived":false,"fork":false,"pushed_at":"2026-03-24T16:48:33.000Z","size":25283,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2026-03-25T20:09:50.808Z","etag":null,"topics":["agent-memory","ai-agent","android","automation","claude-code","codex","desktop-app","developer-tools","electron","mobile","multi-agent","orchestration","react","typescript"],"latest_commit_sha":null,"homepage":"","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"agpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Kujirafu.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2026-03-03T10:00:07.000Z","updated_at":"2026-03-24T16:48:36.000Z","dependencies_parsed_at":"2026-03-06T00:02:48.335Z","dependency_job_id":null,"html_url":"https://github.com/Kujirafu/AgentRune","commit_stats":null,"previous_names":["kujirafu/agentrune"],"tags_count":36,"template":false,"template_full_name":null,"purl":"pkg:github/Kujirafu/AgentRune","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Kujirafu%2FAgentRune","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Kujirafu%2FAgentRune/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Kujirafu%2FAgentRune/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Kujirafu%2FAgentRune/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Kujirafu","download_url":"https://codeload.github.com/Kujirafu/AgentRune/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Kujirafu%2FAgentRune/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31307462,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-02T12:59:32.332Z","status":"ssl_error","status_checked_at":"2026-04-02T12:54:48.875Z","response_time":89,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-memory","ai-agent","android","automation","claude-code","codex","desktop-app","developer-tools","electron","mobile","multi-agent","orchestration","react","typescript"],"created_at":"2026-03-05T23:54:06.226Z","updated_at":"2026-04-02T14:02:13.219Z","avatar_url":"https://github.com/Kujirafu.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# AgentRune\n\nAI agent mobile control panel. Launch, monitor, and schedule AI agents from your phone.\n\n## What is AgentRune?\n\nAgentRune lets you control AI coding agents (Claude Code, Codex, Gemini, Aider, Cline, Cursor) from an Android app. A CLI daemon runs on your PC and exposes a WebSocket + HTTP API through a Cloudflare tunnel, so you can manage sessions from anywhere.\n\n## Features\n\n### Session Control\n- **Session management** -- start, stop, resume, and monitor agent sessions in real time\n- **Session recovery** -- daemon crashes or network drops? One tap to restore full context\n- **Voice input** -- speak prompts to your agents via Android speech recognition\n- **Multi-agent support** -- Claude Code, Codex, Gemini, Aider, Cline, Cursor\n- **Dual-daemon failover** -- dev + release daemons with automatic failover/failback\n- **Settings sync** -- planMode, autoEdit, effort level sync directly to CLI flags\n\n### Automation\n- **Automation scheduling** -- cron-style scheduling with 24+ built-in templates\n- **Skill chains** -- multi-step automation pipelines, chain tasks sequentially or in parallel\n- **Crew system** -- multi-role workflows (PM → engineer → QA) with phase gates and per-role token budgets, built on top of skill chains\n- **Sandbox system** -- 4-tier configurable sandbox (strict / moderate / permissive / none) with prompt conflict scanning\n\n### Trust \u0026 Security\n- **Trust profiles** -- 4 security levels (Autonomous / Supervised / Guarded / Custom) to control what agents can do unattended\n- **Plan panel** -- agents submit execution plans for review before they build\n- **Runtime authority** -- restrict agent operations (file write, HTTP requests, git push) per schedule\n- **Audit log** -- all agent operations recorded for review\n- **AES-256-GCM encryption** for stored secrets and auth tokens\n- **OAuth login** -- GitHub and Google authentication\n\n### Integrations\n- **AgentLore MCP** -- agents query an AI-verified knowledge base\n- **Push notifications** -- FCM alerts for agent decisions, automation completions, version updates\n- **Telemetry** -- self-hosted analytics with opt-out support\n\n## Architecture\n\n```\nPhone (Android app)\n  |\n  | WebSocket + HTTPS (Cloudflare tunnel)\n  |\nPC (CLI daemon)\n  |\n  |-- PTY adapter (real-time terminal I/O)\n  |-- JSONL watcher (session replay / token tracking)\n  |-- Express HTTP API\n  |-- AgentLore MCP server\n```\n\n**Monorepo structure:**\n\n| Directory | Description |\n|-----------|-------------|\n| `app/` | Vite + React + TypeScript + Capacitor (Android) |\n| `cli/` | Node.js CLI daemon (Express + WebSocket + PTY) |\n| `docs/` | Design documents and plans |\n\n## Getting Started\n\n### Prerequisites\n\n- Node.js 20+\n- Android SDK (for APK builds)\n- One or more AI agents installed (e.g. `claude`, `codex`, `cursor`)\n\n### Install\n\n```bash\nnpm install\n```\n\n### Run the CLI daemon (dev)\n\n```bash\ncd cli \u0026\u0026 npm run dev\n```\n\nThis starts the daemon on port 3457 in foreground mode with file watching.\n\n### Build the app\n\n```bash\n# Development build\ncd app \u0026\u0026 npm run build:dev\n\n# Production build\ncd app \u0026\u0026 npm run build\n```\n\n### Build APK\n\n```bash\ncd app\nnpx cap sync android\ncd android \u0026\u0026 ./gradlew assembleDebug\n```\n\nAPK output: `app/android/app/build/outputs/apk/debug/app-debug.apk`\n\n### Install the CLI globally\n\n```bash\nnpm install -g agentrune\nagentrune start\n```\n\n## Community\n\nJoin us for testing, feedback, and discussion:\n\n- **Telegram** -- [AgentLore \u0026 AgentRune](https://t.me/AgentLore_n_AgentRune)\n- **X (Twitter)** -- [@AGLO_Official](https://x.com/AGLO_Official)\n- **Moltbook** -- [agentlore](https://www.moltbook.com/u/agentlore)\n\n## Project Memory\n\nThis repository keeps public project documentation in tracked files such as `README.md`, `docs/`, source comments, and tests.\n\nAgent workflow memory lives separately in local `.agentrune/` files and is not intended to be the public documentation layer for the AGPL repository. See [docs/project-memory-policy.md](docs/project-memory-policy.md).\n\nThe local memory system is index-first:\n- `.agentrune/agentlore.md` is the short index\n- `.agentrune/context/*.md` stores detailed sections\n- Agents should route to the relevant section first, then read only what they need\n\nThe CLI and bundled desktop install now expose this system directly:\n\n```bash\nagentrune memory init\nagentrune memory sections\nagentrune memory route \"Fix tunnel auth bug in ws-server\" --file cli/src/server/ws-server.ts\nagentrune memory search \"tunnel auth\"\nagentrune mcp\n```\n\n`agentrune mcp` memory tools operate on the current working directory, or `AGENTRUNE_PROJECT_CWD` when that environment variable is set.\n\n## License\n\n[AGPL-3.0](https://www.gnu.org/licenses/agpl-3.0.html)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkujirafu%2Fagentrune","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkujirafu%2Fagentrune","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkujirafu%2Fagentrune/lists"}