{"id":51719728,"url":"https://github.com/AutoResearch-Factory/Agon","last_synced_at":"2026-07-18T12:00:45.241Z","repository":{"id":369355548,"uuid":"1273471983","full_name":"AutoResearch-Factory/Agon","owner":"AutoResearch-Factory","description":"Autonomous Large-Scale Minimal Omnidisciplinary Research System","archived":false,"fork":false,"pushed_at":"2026-07-17T05:02:09.000Z","size":270,"stargazers_count":25,"open_issues_count":1,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-07-17T07:05:27.026Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/AutoResearch-Factory.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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":null,"dco":null,"cla":null}},"created_at":"2026-06-18T14:56:15.000Z","updated_at":"2026-07-17T05:02:13.000Z","dependencies_parsed_at":null,"dependency_job_id":"47dbb1fc-979b-4c03-993b-ee1d3afbda56","html_url":"https://github.com/AutoResearch-Factory/Agon","commit_stats":null,"previous_names":["autoresearch-factory/agon"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/AutoResearch-Factory/Agon","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AutoResearch-Factory%2FAgon","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AutoResearch-Factory%2FAgon/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AutoResearch-Factory%2FAgon/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AutoResearch-Factory%2FAgon/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/AutoResearch-Factory","download_url":"https://codeload.github.com/AutoResearch-Factory/Agon/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AutoResearch-Factory%2FAgon/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35616739,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-18T02:00:07.223Z","response_time":61,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":[],"created_at":"2026-07-17T09:00:31.378Z","updated_at":"2026-07-18T12:00:45.229Z","avatar_url":"https://github.com/AutoResearch-Factory.png","language":"Python","funding_links":[],"categories":["Open-Source Implementations","Full list","🔬 Research-agent systems","Source Catalog","Harnesses \u0026 orchestration","🚀 Specialized Agents","Multi-Agent Swarms","🔬 Autonomous Research \u0026 Self-Improving Agents"],"sub_categories":["Scientific Research","Orchestrators \u0026 autonomous loops","🔬 Research Agents","Research-Agent Systems"],"readme":"# Agon\n\nEnglish | [中文](README_zh.md)\n\nAgon ([paper](https://arxiv.org/abs/2606.24177)) is an automated research system. It keeps the workflow minimal and explicit: `topic -\u003e idea -\u003e proposal -\u003e experiment`.\n\n![Agon workflow](figures/figure_xp.png)\n\n## Quick start\n\nClone [Agon](https://github.com/AutoResearch-Factory/Agon) and [agon-artifacts](https://github.com/AutoResearch-Factory/agon-artifacts):\n\n```\ngit clone https://github.com/AutoResearch-Factory/Agon.git\ngit clone https://github.com/AutoResearch-Factory/agon-artifacts.git\n```\n\nPut the two directories side by side:\n\n```\n.\n├── Agon/\n└── agon-artifacts/\n```\n\nThen run Claude Code from the artifacts repository:\n\n```\ncd agon-artifacts\nclaude --plugin-dir ../Agon --dangerously-skip-permissions --model claude-sonnet-5[1m]\n```\n\nIn Claude Code, use these commands to move the research forward:\n\n- `/idea-tick`: create, review, refine, and literature-check ideas for a topic.\n- `/proposal-tick`: turn selected ideas into reviewed proposals.\n- `/experiment-tick`: coordinate scientist, coder, auditor, and reviewer roles for one workspace.\n- `/deep-lit-tick`: run the shared deep literature loop used by the other stages.\n\n## Example Prompts\n\n```\n/deep-lit-tick Exhaustively survey the literature on \u003ctopic\u003e, and write the result to topics/mmdd-\u003cslug\u003e-landscape.md.\n/idea-tick \u003ctopic-slug\u003e \u003ctopic\u003e is becoming important. Brainstorm several research ideas.\n/idea-tick \u003cidea-slug\u003e I have a vague idea about \u003ctopic\u003e. Create the topic file, create the idea file, and refine the idea.\n/proposal-tick \u003cidea-1\u003e \u003cidea-2\u003e \u003cidea-3\u003e Generate proposals for these ideas.\n/experiment-tick \u003cslug\u003e Start the experiment.\n/experiment-tick \u003cslug\u003e This is a debugging run. First explain the full procedure, then pause for my approval after each agent call.\n```\n\n## Layout\n\nAgon itself is a Claude Code plugin. Run it from a separate data workspace, commonly named `agon-artifacts`, so prompts/code and research data can evolve independently.\n\nExpected data workspace layout:\n\n```\nagon-artifacts/\n├── topics/\n├── ideas/\n└── workspace/\n```\n\nOptional local settings live at `.settings.toml`. Start from `.settings.example.toml` when you need to customize model routing or parallelism.\n\n## claude-ds installation\n\n`claude-ds` is DeepSeek-backed Claude Code.\n\n1. Before first launch, create an empty `~/.claude-ds/` and only handle symlinks. Claude will generate the rest on first launch:\n   - Settings side (symlink): `CLAUDE.md` `mcp-needs-auth-cache.json` `memory/` `plugins/` `settings.json` `settings.local.json` `skills/`\n   - Session side (isolated): `backups/` `cache/` `downloads/` `ide/` `stats-cache.json` `projects/` `sessions/` `session-env/` `file-history/` `history.jsonl` `paste-cache/` `shell-snapshots/` `.claude.json`\n   - The symlinks let `claude` and `claude-ds` share skills, MCP, and plugins\n\n2. Add the `claude-ds()` function to `~/.bashrc`:\n```\nclaude-ds() {\n  CLAUDE_CONFIG_DIR=\"$HOME/.claude-ds\" \\\n  ANTHROPIC_BASE_URL=\"https://api.deepseek.com/anthropic\" \\\n  ANTHROPIC_AUTH_TOKEN=\"$DEEPSEEK_API_KEY\" \\\n  ANTHROPIC_MODEL=\"deepseek-v4-pro[1m]\" \\\n  ANTHROPIC_SMALL_FAST_MODEL=\"deepseek-v4-flash\" \\\n  claude --effort max \"$@\"\n}\n```\nNotes:\n- Put it before the `# If not running interactively` guard, so non-interactive shells such as Claude Code's Bash tool can see it\n- Add the `[1m]` suffix to the model ID to unlock 1M context\n3. Run `claude-ds`\n\n## Required Claude Code settings\n\n| Item | Purpose |\n|------|------|\n| `DISABLE_TELEMETRY` | Disable Statsig telemetry (usage stats, no code or file paths) |\n| `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS` | Let the main agent resume background/asynchronous subagents, and let subagents message each other |\n| `cleanupPeriodDays: 3650` | Keep session history under `~/.claude/projects/` (default cleanup is after 30 days); set this at the top level of `~/.claude/settings.json` |\n\n**Configure statusline**\n\nTell `claude` or `claude-ds`: \"call statusline-setup, I want `[5h:6% 7d:69%(2d17h)] Ctx:7% Opus 4.6 (1M context)`\"\n\n## References\n\n- [ARIS](https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep)\n- [AutoResearch-SibylSystem](https://github.com/Sibyl-Research-Team/AutoResearch-SibylSystem)\n- [Telegram MCP Communicator](https://github.com/WhymustIhaveaname/mcp-communicator-telegram)\n- [Claude Memory Manager](https://github.com/WhymustIhaveaname/claude-memory-manager)\n- [humanizer](https://github.com/blader/humanizer)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FAutoResearch-Factory%2FAgon","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FAutoResearch-Factory%2FAgon","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FAutoResearch-Factory%2FAgon/lists"}