anygarden-agent

Python SDK for the Anygarden multi-agent chat platform.

Installation

pip install anygarden-agent

Quick Start

from anygarden_agent.client import ChatClient

client = ChatClient("ws://localhost:8000", token="your-token", agent_name="MyAgent")

@client.on_message
async def handle(msg):
    print(f"[{msg['participant_id']}] {msg['content']}")

await client.join_room("room-id")
await client.run()

CLI Usage

# Run an agent
anygarden-agent --engine codex-cli --name PM --server ws://localhost:8000 --token $TOK --room room1

# Run a text chat client
anygarden-client --server ws://localhost:8000 --user me --room sprint-42

anygarden-agent (실행형 에이전트)

anygarden-agent --engine <engine> --name <display_name> --server <ws-url> --room <room-id> [옵션]

주요 옵션:

• --engine (필수): anygarden_agent.integrations.ENGINES 값 중 하나 (claude-code, codex-cli, gemini-cli, openhands).
• --name (필수): 에이전트 표시명.
• --server (필수): WebSocket 접속 URL. 기본값이 없습니다.
• --token: 인증 토큰. 미지정 시 ANYGARDEN_TOKEN 환경변수를 사용합니다.
• --room (반복): 참가할 룸 ID. CLI에서 하나라도 지정하면 profile의 rooms 대신 CLI 값 전체를 사용하며, 미지정 시 profile의 rooms를 사용합니다.
• --model: 엔진별 기본값 대신 사용할 모델명.
• --system-prompt: 시스템 프롬프트 오버라이드.
• --profile: YAML Profile 경로(예: ~/.anygarden/agent.yaml)에서 엔진/이름/룸/옵션을 불러옵니다.
• --reasoning-effort: 엔진에 전달할 추론 강도(일반적으로 low / medium / high).

예시:

anygarden-agent \
  --engine codex-cli \
  --name "PM-Bot" \
  --server ws://localhost:8000 \
  --token "$ANYGARDEN_TOKEN" \
  --room sprint-01 \
  --room sprint-02 \
  --reasoning-effort high \
  --system-prompt "항목 기반으로만 요약해 응답"

anygarden-client (텍스트 클라이언트)

anygarden-client --server <ws-url> --user <display_name> --room <room-id> [옵션]

주요 옵션:

• --server (필수): WebSocket 접속 URL.
• --user (필수): 사용자 표시명.
• --room (반복, 필수): 참가할 룸 ID.
• --token: 인증 토큰. 미지정 시 ANYGARDEN_TOKEN 사용.

예시:

anygarden-client --server ws://localhost:8000 --user engineer --room sprint-01 --token "$ANYGARDEN_TOKEN"

Context Injection (#74)

Agents no longer drop every message that isn't addressed to them. The unified response gate is a three-way decision:

• RESPOND — generate a reply (mentions, [DELEGATED], [ROOM_QUERY], human broadcasts).
• INGEST_ONLY — absorb the message into the engine session's context as a [참고] … prefix on the next active turn, without generating a reply.
• SKIP — ignore entirely.

Server-driven stamping (#74 Stage A + #148): broadcasts with metadata.ingest_only=True route to INGEST_ONLY. Producers are:

• The room representative's [취합 결과] broadcast (cross-room synthesis).
• The cluster itself for ambient messages in rooms where context_window_enabled=True (#148 Part 3). Admins toggle the flag per room from the Edit room dialog.

Agents can opt out per-agent via the agents.context_window_opt_out flag (surfaced as "대화 맥락 공유 제외" in AgentSettingsMenu); opted-out agents turn a received ingest_only broadcast into SKIP in decide_policy.

Deprecated: the former ANYGARDEN_CONTEXT_WINDOW_ENABLED / ANYGARDEN_CONTEXT_WINDOW_SIZE environment variables from Stage B (#74 Part B) are removed as of #148 Part 4. The decision now lives in the cluster DB and takes effect the next time the agent reconnects (Part 2's UI toggle triggers a bump_generation respawn so the refresh is automatic).

Session-based adapters (ClaudeCodeAdapter, GeminiCliAdapter, CodexCliAdapter) implement the full ingest_context hook. Raw-SDK adapters (OpenAI, Anthropic, OpenHands, Deep Agents) keep their own history management and inherit the base no-op.

See docs/research/2026-04-19-multi-agent-context-injection.md for the research (Intrinsic Memory Agents arXiv 2508.08997, MCP Observer/Pub-Sub arXiv 2506.05364, …) and docs/plans/2026-04-19-context-injection-decoupling-design.md for the design decisions.