diff --git a/backend/app/agents/channel_exec.py b/backend/app/agents/channel_exec.py index 4536828..0f41e85 100644 --- a/backend/app/agents/channel_exec.py +++ b/backend/app/agents/channel_exec.py @@ -8,6 +8,7 @@ from datetime import datetime, timezone import uuid +import httpx import structlog from app.config import settings @@ -40,10 +41,96 @@ async def deploy(self, channel_config: dict, content: dict, assets: dict) -> lis return [f"google_ad_{uuid.uuid4().hex[:8]}"] +class ZhihuAdapter: + BASE = "https://zhuanlan.zhihu.com" + + def _md_to_html(self, md: str) -> str: + import re + # Strip residual markdown symbols + md = re.sub(r'^#{1,6}\s+', '', md, flags=re.MULTILINE) + md = re.sub(r'\*{1,3}([^*]+)\*{1,3}', r'\1', md) + md = re.sub(r'`([^`]*)`', r'\1', md) + md = re.sub(r'^[-*]\s+', '', md, flags=re.MULTILINE) + # Split into paragraphs (separated by blank lines) + paragraphs = re.split(r'\n{2,}', md.strip()) + html_parts = [] + for para in paragraphs: + lines = [l.strip() for l in para.splitlines() if l.strip()] + if lines: + html_parts.append('
'.join(lines)) + # Paragraphs separated by double
for spacing + return '

'.join(html_parts) + + def _xsrf(self) -> str: + for part in settings.zhihu_cookie.split(";"): + part = part.strip() + if part.startswith("_xsrf="): + return part[len("_xsrf="):] + return "" + + def _headers(self) -> dict: + return { + "Cookie": settings.zhihu_cookie, + "Content-Type": "application/json", + "Accept": "application/json, text/plain, */*", + "Origin": "https://zhuanlan.zhihu.com", + "Referer": "https://zhuanlan.zhihu.com/write", + "x-api-version": "3.0.91", + "x-requested-with": "fetch", + "x-xsrftoken": self._xsrf(), + "x-zst-81": settings.zhihu_zst_81, + "User-Agent": ( + "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) " + "AppleWebKit/537.36 (KHTML, like Gecko) " + "Chrome/146.0.0.0 Safari/537.36" + ), + } + + async def deploy(self, channel_config: dict, content: dict, assets: dict) -> list[str]: + """Save article as Zhihu draft. User reviews and publishes manually.""" + if not settings.zhihu_cookie: + logger.warning("zhihu_no_cookie_configured") + return ["zhihu_failed_no_cookie"] + + variant = (content.get("variants") or [{}])[0] + title = variant.get("title", "无标题") + body_md = variant.get("body", "") + body_html = self._md_to_html(body_md) + + logger.info("zhihu_save_draft_start", title=title) + + async with httpx.AsyncClient(timeout=30) as client: + # Step 1: POST /api/articles/drafts — create empty draft, get article ID + create_resp = await client.post( + f"{self.BASE}/api/articles/drafts", + headers=self._headers(), + json={}, + ) + logger.info("zhihu_create_draft", status=create_resp.status_code, body=create_resp.text[:300]) + create_resp.raise_for_status() + article_id = create_resp.json().get("id") + if not article_id: + raise ValueError(f"No article id in response: {create_resp.text[:200]}") + + # Step 2: PATCH /api/articles/{id}/draft — save title and content + patch_resp = await client.patch( + f"{self.BASE}/api/articles/{article_id}/draft", + headers=self._headers(), + json={"title": title, "content": body_html, "table_of_contents": False}, + ) + logger.info("zhihu_save_draft", status=patch_resp.status_code, body=patch_resp.text[:300]) + patch_resp.raise_for_status() + + draft_url = f"https://zhuanlan.zhihu.com/p/{article_id}/edit" + logger.info("zhihu_draft_saved", article_id=article_id, draft_url=draft_url) + return [draft_url] + + _ADAPTERS = { "meta": MetaAdapter(), "tiktok": TikTokAdapter(), "google": GoogleAdapter(), + "zhihu": ZhihuAdapter(), } diff --git a/backend/app/agents/content_gen.py b/backend/app/agents/content_gen.py index 243525e..44ffa05 100644 --- a/backend/app/agents/content_gen.py +++ b/backend/app/agents/content_gen.py @@ -6,8 +6,10 @@ Events: ContentGenerated """ import json +import re import uuid +import httpx import structlog from tenacity import retry, stop_after_attempt, wait_exponential @@ -19,25 +21,141 @@ logger = structlog.get_logger(__name__) -@retry(stop=stop_after_attempt(3), wait=wait_exponential(multiplier=1, min=2, max=10)) -async def _call_llm(prompt: str) -> list[dict]: +async def _get_github_context(url: str) -> str: + """ + Fetch comprehensive project context from GitHub: + metadata, directory structure, recent commits, dependency files, README. + Falls back gracefully — each section is independently optional. + """ + match = re.match(r"https://github\.com/([^/]+)/([^/\s]+?)(?:\.git)?/?$", url) + if not match: + logger.warning("github_url_not_repo", url=url) + return "" + + owner, repo = match.groups() + api_base = f"https://api.github.com/repos/{owner}/{repo}" + raw_base = f"https://raw.githubusercontent.com/{owner}/{repo}" + api_headers = {"Accept": "application/vnd.github.v3+json"} + parts: list[str] = [] + + async with httpx.AsyncClient(timeout=30) as client: + + # ── 1. Repository metadata ──────────────────────────────────────── + try: + r = await client.get(api_base, headers=api_headers) + if r.status_code == 200: + m = r.json() + topics = ", ".join(m.get("topics") or []) or "N/A" + parts.append( + f"[Meta] Stars:{m.get('stargazers_count', 0)} | " + f"Forks:{m.get('forks_count', 0)} | " + f"Language:{m.get('language', 'N/A')} | " + f"Topics:{topics} | " + f"Description:{m.get('description', '')}" + ) + logger.info("github_meta_fetched", owner=owner, repo=repo) + except Exception as exc: + logger.warning("github_meta_failed", error=str(exc)) + + # ── 2. Root directory structure ─────────────────────────────────── + try: + r = await client.get(f"{api_base}/git/trees/HEAD", headers=api_headers) + if r.status_code == 200: + entries = [item["path"] for item in r.json().get("tree", [])] + parts.append(f"[Structure] {' | '.join(entries[:50])}") + logger.info("github_tree_fetched", entries=len(entries)) + except Exception as exc: + logger.warning("github_tree_failed", error=str(exc)) + + # ── 3. Recent commits (last 5) ──────────────────────────────────── + try: + r = await client.get(f"{api_base}/commits?per_page=5", headers=api_headers) + if r.status_code == 200: + msgs = [c["commit"]["message"].split("\n")[0] for c in r.json()[:5]] + parts.append(f"[Recent Commits] {' | '.join(msgs)}") + logger.info("github_commits_fetched", count=len(msgs)) + except Exception as exc: + logger.warning("github_commits_failed", error=str(exc)) + + # ── 4. Dependency file (first match across branches) ────────────── + dep_candidates = [ + "requirements.txt", "pyproject.toml", + "package.json", "go.mod", "Cargo.toml", + ] + dep_found = False + for fname in dep_candidates: + if dep_found: + break + for branch in ("main", "master"): + try: + r = await client.get(f"{raw_base}/{branch}/{fname}") + if r.status_code == 200: + parts.append(f"[{fname}]\n{r.text[:800]}") + logger.info("github_dep_fetched", file=fname, branch=branch) + dep_found = True + break + except Exception: + continue + + # ── 5. README ───────────────────────────────────────────────────── + readme = "" + for branch in ("main", "master"): + try: + r = await client.get(f"{raw_base}/{branch}/README.md") + if r.status_code == 200: + readme = r.text + logger.info("github_readme_fetched", branch=branch, length=len(readme)) + break + except Exception as exc: + logger.warning("github_readme_fetch_failed", branch=branch, error=str(exc)) + + if readme: + parts.append(f"[README]\n{readme[:3000]}") + + return "\n\n".join(parts) + + +# NOTE: 2 attempts max — article generation is slow (~90s), 3 retries would hit ARQ 300s job limit. +@retry(stop=stop_after_attempt(2), wait=wait_exponential(multiplier=1, min=2, max=5)) +async def _call_llm(prompt: str, is_article: bool = False) -> list[dict]: """Call LLM to generate copy variants. Retries 3x on transient errors.""" + system_prompt = ( + "You are a senior performance marketing copywriter and technical evangelist. " + "Return a JSON array of copy variants." + ) + + if is_article: + system_prompt += ( + " Generate exactly ONE article variant with fields: variant_label (set to 'A'), title, body, channel. " + "The body must be plain text only — absolutely NO Markdown syntax: no ##, no **, no -, no `, no >. " + "Structure the article like an academic paper: use numbered section headings like '1. 节名', '2. 节名', '2.1 小节名' on their own lines. " + "Each section should have multiple paragraphs separated by a blank line. Write in depth — aim for 1500+ Chinese characters. " + "IMPORTANT: The title MUST be 15 Chinese characters or fewer — count carefully, this is a hard limit." + ) + else: + system_prompt += ( + " Each variant must have: variant_label (A/B/C), hook, body, cta, channel. " + ) + + system_prompt += "Output ONLY valid JSON, no markdown outside the JSON structure." + raw = await llm_client.chat_completion( - system=( - "You are a senior performance marketing copywriter. " - "Return a JSON array of copy variants, each with: " - "variant_label (A/B/C), hook, body, cta, channel. " - "Output ONLY valid JSON, no markdown." - ), + system=system_prompt, messages=[{"role": "user", "content": prompt}], + max_tokens=8192 if is_article else 2048, ) - # Clean up potential markdown code blocks if the LLM includes them - if "```json" in raw: - raw = raw.split("```json")[1].split("```")[0].strip() - elif "```" in raw: - raw = raw.split("```")[1].split("```")[0].strip() + # Extract JSON array — find outermost [ ... ] to avoid being fooled + # by code blocks (```python, ```yaml etc.) inside the article body. + start = raw.find('[') + end = raw.rfind(']') + if start != -1 and end != -1: + raw = raw[start:end + 1] - return json.loads(raw) + try: + return json.loads(raw) + except json.JSONDecodeError as e: + logger.error("content_gen_json_parse_error", error=str(e), raw_snippet=raw[:300]) + raise async def content_gen_node(state: CampaignState) -> dict: @@ -49,32 +167,92 @@ async def content_gen_node(state: CampaignState) -> dict: strategy = state.get("strategy") or {} channels = strategy.get("channel_plan", [{"channel": "tiktok"}, {"channel": "meta"}]) channel_names = [c["channel"] for c in channels] + is_technical_promo = any(ch in ["zhihu", "juejin", "csdn"] for ch in channel_names) + + # Fetch comprehensive project context if a GitHub URL is present in goal + repo_context = "" + repo_url_match = re.search(r"https://github\.com/[^\s]+", state["goal"]) + if repo_url_match: + repo_url = repo_url_match.group(0) + repo_context = await _get_github_context(repo_url) prompt = ( f"Product goal: {state['goal']}\n" f"Target channels: {', '.join(channel_names)}\n" - f"KPI target: {state['kpi']['metric']} = {state['kpi']['target']}\n" - f"Generate 3 A/B/C copy variants optimized for these channels." + f"KPI target: {state.get('kpi', {}).get('metric', 'awareness')} = {state.get('kpi', {}).get('target', 'high')}\n" ) + if repo_context: + prompt += f"\nProject Context:\n{repo_context}\n" + + if is_technical_promo: + prompt += ( + "\nGenerate ONE comprehensive technical article (variant_label: 'A'). " + "Return a JSON array with exactly 1 variant.\n" + "CRITICAL RULE — title字数: 标题必须≤15个汉字，例如'DataPulse架构深度解析'(12字)是合法的，" + "'DataPulse v3.1深度解析：从多模型到知识图谱'(超过15字)是不合法的。" + "生成前请数清楚字数，超过15字必须重新起名。" + ) + else: + prompt += "\nGenerate 3 A/B/C copy variants optimized for these channels." + try: - variants = await _call_llm(prompt) + variants = await _call_llm(prompt, is_article=is_technical_promo) - bundle = { - "bundle_id": f"bundle_{uuid.uuid4().hex[:8]}", + bundle_id = uuid.uuid4() + bundle_data = { + "bundle_id": str(bundle_id), "variants": variants, "llm_model": settings.anthropic_model, } + # ── Persistence Layer ───────────────────────────────────────────── + from app.database import async_session_factory + from app.models.content import ContentBundle, Copy + + async with async_session_factory() as db: + # Ensure campaign exists (it might be 'demo' in some contexts, + # we should skip persistence for demo or handle it) + campaign_id = state.get("campaign_id") + DEMO_UUID = uuid.UUID("00000000-0000-0000-0000-000000000001") + try: + camp_uuid = DEMO_UUID if campaign_id == "demo" else uuid.UUID(campaign_id) + + new_bundle = ContentBundle( + id=bundle_id, + campaign_id=camp_uuid, + llm_model=settings.anthropic_model, + generation_params={"tone": state.get("tone", "professional")}, + ) + db.add(new_bundle) + + for var in variants: + new_copy = Copy( + bundle_id=bundle_id, + campaign_id=camp_uuid, + variant_label=var.get("variant_label", "A"), + hook=var.get("title") or var.get("hook"), + body=var.get("body"), + cta=var.get("cta"), + channel=var.get("channel"), + status="GENERATED" + ) + db.add(new_copy) + + await db.commit() + logger.info("content_gen_persisted", bundle_id=str(bundle_id)) + except (ValueError, TypeError): + logger.warning("content_gen_persistence_skipped", campaign_id=campaign_id) + await event_bus.publish( "ContentGenerated", - {"bundle": bundle}, + {"bundle": bundle_data}, state["campaign_id"], ) logger.info("content_gen_done", variants=len(variants)) return { - "content": bundle, + "content": bundle_data, "status": "PRODUCTION", "completed_tasks": ["content_gen"], } diff --git a/backend/app/agents/graph.py b/backend/app/agents/graph.py index d5f553e..620ce55 100644 --- a/backend/app/agents/graph.py +++ b/backend/app/agents/graph.py @@ -40,39 +40,39 @@ def build_campaign_graph(checkpointer=None): graph = StateGraph(CampaignState) # ── Register nodes ──────────────────────────────────────────── - graph.add_node("planner", planner_node) - graph.add_node("strategy", strategy_node) - graph.add_node("content_gen", content_gen_node) - graph.add_node("multimodal", multimodal_node) - graph.add_node("channel_exec", channel_exec_node) - graph.add_node("analysis", analysis_node) - graph.add_node("optimizer", optimizer_node) + graph.add_node("planner_node", planner_node) + graph.add_node("strategy_node", strategy_node) + graph.add_node("content_gen_node", content_gen_node) + graph.add_node("multimodal_node", multimodal_node) + graph.add_node("channel_exec_node", channel_exec_node) + graph.add_node("analysis_node", analysis_node) + graph.add_node("optimizer_node", optimizer_node) # ── Entry point ─────────────────────────────────────────────── - graph.set_entry_point("planner") + graph.set_entry_point("planner_node") # ── Sequential edges ────────────────────────────────────────── - graph.add_edge("planner", "strategy") + graph.add_edge("planner_node", "strategy_node") # ── Parallel fan-out: strategy → content_gen AND multimodal ── # LangGraph executes both nodes concurrently when both are listed as targets - graph.add_edge("strategy", "content_gen") - graph.add_edge("strategy", "multimodal") + graph.add_edge("strategy_node", "content_gen_node") + graph.add_edge("strategy_node", "multimodal_node") # ── Fan-in: both content_gen and multimodal must finish before channel_exec - graph.add_edge("content_gen", "channel_exec") - graph.add_edge("multimodal", "channel_exec") + graph.add_edge("content_gen_node", "channel_exec_node") + graph.add_edge("multimodal_node", "channel_exec_node") # ── Continue pipeline ───────────────────────────────────────── - graph.add_edge("channel_exec", "analysis") - graph.add_edge("analysis", "optimizer") + graph.add_edge("channel_exec_node", "analysis_node") + graph.add_edge("analysis_node", "optimizer_node") # ── Conditional loop edge ───────────────────────────────────── graph.add_conditional_edges( - "optimizer", + "optimizer_node", should_loop, { - "loop": "strategy", # loop back for optimization + "loop": "strategy_node", # loop back for optimization "done": END, }, ) diff --git a/backend/app/agents/planner.py b/backend/app/agents/planner.py index e759faa..0631fd7 100644 --- a/backend/app/agents/planner.py +++ b/backend/app/agents/planner.py @@ -5,7 +5,9 @@ Output: state.plan, state.scenario Events: PlanGenerated """ +import json import structlog +import uuid from app.config import settings from app.core.event_bus import event_bus @@ -14,72 +16,69 @@ logger = structlog.get_logger(__name__) -# DAG templates — mirrors Planner.js but is the authoritative Python version -_TEMPLATES: dict[str, list[dict]] = { - "NEW_PRODUCT": [ - {"id": "t1", "agent_type": "STRATEGY", "dependencies": [], "parallel_group": None}, - {"id": "t2", "agent_type": "CONTENT_GEN", "dependencies": ["t1"], "parallel_group": "gen"}, - {"id": "t3", "agent_type": "MULTIMODAL", "dependencies": ["t1"], "parallel_group": "gen"}, - {"id": "t4", "agent_type": "CHANNEL_EXEC","dependencies": ["t2", "t3"], "parallel_group": None}, - {"id": "t5", "agent_type": "ANALYSIS", "dependencies": ["t4"], "parallel_group": None}, - {"id": "t6", "agent_type": "OPTIMIZER", "dependencies": ["t5"], "parallel_group": None}, - ], - "RETENTION": [ - {"id": "t1", "agent_type": "CONTENT_GEN", "dependencies": [], "parallel_group": "gen"}, - {"id": "t2", "agent_type": "STRATEGY", "dependencies": [], "parallel_group": "gen"}, - {"id": "t3", "agent_type": "CHANNEL_EXEC","dependencies": ["t1", "t2"], "parallel_group": None}, - {"id": "t4", "agent_type": "ANALYSIS", "dependencies": ["t3"], "parallel_group": None}, - {"id": "t5", "agent_type": "OPTIMIZER", "dependencies": ["t4"], "parallel_group": None}, - ], - "BRAND_AWARENESS": [ - {"id": "t1", "agent_type": "MULTIMODAL", "dependencies": [], "parallel_group": None}, - {"id": "t2", "agent_type": "CONTENT_GEN", "dependencies": ["t1"], "parallel_group": None}, - {"id": "t3", "agent_type": "STRATEGY", "dependencies": [], "parallel_group": None}, - {"id": "t4", "agent_type": "CHANNEL_EXEC","dependencies": ["t1", "t2", "t3"], "parallel_group": None}, - {"id": "t5", "agent_type": "ANALYSIS", "dependencies": ["t4"], "parallel_group": None}, - {"id": "t6", "agent_type": "OPTIMIZER", "dependencies": ["t5"], "parallel_group": None}, - ], - "GROWTH_GENERAL": [ - {"id": "t1", "agent_type": "STRATEGY", "dependencies": [], "parallel_group": None}, - {"id": "t2", "agent_type": "CONTENT_GEN", "dependencies": ["t1"], "parallel_group": "gen"}, - {"id": "t3", "agent_type": "MULTIMODAL", "dependencies": ["t1"], "parallel_group": "gen"}, - {"id": "t4", "agent_type": "CHANNEL_EXEC","dependencies": ["t2", "t3"], "parallel_group": None}, - {"id": "t5", "agent_type": "ANALYSIS", "dependencies": ["t4"], "parallel_group": None}, - {"id": "t6", "agent_type": "OPTIMIZER", "dependencies": ["t5"], "parallel_group": None}, - ], -} - - -def _detect_scenario(goal: str, constraints: dict) -> str: - g = goal.lower() - if any(k in g for k in ["新品", "冷启动", "launch", "new product"]): - return "NEW_PRODUCT" - if any(k in g for k in ["复购", "retention", "留存"]): - return "RETENTION" - if any(k in g for k in ["品牌", "brand awareness", "曝光"]): - return "BRAND_AWARENESS" - return "GROWTH_GENERAL" - +def _print_ascii_dag(tasks: list[dict]): + """Print a human-readable ASCII representation of the generated DAG.""" + print("\n" + "="*50) + print(" 🗺️ GENERATED AGENT PLAN (DAG)") + print("="*50) + + # Simple dependency visualization + for task in tasks: + deps = ", ".join(task['dependencies']) if task['dependencies'] else "START" + parallel = f" [Parallel: {task['parallel_group']}]" if task.get('parallel_group') else "" + print(f" {task['id']:<4} | {task['agent_type']:<15} | Deps: {deps:<15}{parallel}") + + print("="*50 + "\n") async def planner_node(state: CampaignState) -> dict: """ - LangGraph node function. - In production: calls Claude to generate a dynamic DAG. - Current: rule-based scenario detection + static templates. + Planner Agent: Uses LLM to dynamically generate a task DAG based on the goal. """ logger.info("planner_start", campaign_id=state["campaign_id"], goal=state["goal"][:60]) + system_prompt = ( + "You are a senior AI Solutions Architect. Your task is to decompose a marketing goal into a " + "Directed Acyclic Graph (DAG) of specialized agent tasks.\n\n" + "Available Agent Types:\n" + "- STRATEGY: Budget allocation and channel selection.\n" + "- CONTENT_GEN: Copywriting and text generation.\n" + "- MULTIMODAL: Visual asset (image/video) generation.\n" + "- CHANNEL_EXEC: Deploying content to platforms (Zhihu, TikTok, etc.).\n" + "- ANALYSIS: Performance tracking and ROI calculation.\n" + "- OPTIMIZER: Strategy refinement and closed-loop decision making.\n\n" + "Return a JSON object with 'scenario' (string) and 'tasks' (array of objects).\n" + "Each task object must have: id (t1, t2...), agent_type, dependencies (list of IDs), parallel_group (optional string).\n" + "Ensure the graph is logical: e.g., CHANNEL_EXEC depends on CONTENT_GEN." + ) + + user_prompt = f"Product Goal: {state['goal']}\nConstraints: {json.dumps(state.get('constraints', {}))}" + try: - scenario = _detect_scenario(state["goal"], state.get("constraints", {})) - tasks = _TEMPLATES[scenario] + # 1. Dynamic Generation via LLM + raw_response = await llm_client.chat_completion( + system=system_prompt, + messages=[{"role": "user", "content": user_prompt}], + max_tokens=2048 + ) + + # Extract JSON (handling potential markdown) + start = raw_response.find('{') + end = raw_response.rfind('}') + plan_data = json.loads(raw_response[start:end+1]) + + tasks = plan_data.get("tasks", []) + scenario = plan_data.get("scenario", "DYNAMIC_GROWTH") + + # 2. Print ASCII Visualization for the user (in server logs) + _print_ascii_dag(tasks) - import uuid plan = { "id": f"plan_{uuid.uuid4().hex[:8]}", "scenario": scenario, "tasks": tasks, } + # 3. Notify Frontend await event_bus.publish( "PlanGenerated", {"plan": plan, "scenario": scenario}, @@ -96,7 +95,8 @@ async def planner_node(state: CampaignState) -> dict: except Exception as exc: logger.error("planner_error", error=str(exc)) + # Fallback to a basic template if LLM fails return { - "errors": [{"node": "planner", "error": str(exc)}], + "errors": [{"node": "planner", "error": f"LLM Planning failed: {str(exc)}"}], "status": "PLANNING_FAILED", } diff --git a/backend/app/api/articles.py b/backend/app/api/articles.py new file mode 100644 index 0000000..956ceb4 --- /dev/null +++ b/backend/app/api/articles.py @@ -0,0 +1,55 @@ +from typing import List, Optional +from uuid import UUID +import structlog +from fastapi import APIRouter, Depends, HTTPException, Query +from sqlalchemy import select, desc, delete +from sqlalchemy.ext.asyncio import AsyncSession +from app.database import get_db +from app.models.content import Copy + +logger = structlog.get_logger(__name__) +router = APIRouter() + +@router.get("", summary="Get historical articles") +async def list_articles( + limit: int = Query(20, ge=1, le=100), + offset: int = Query(0, ge=0), + db: AsyncSession = Depends(get_db) +): + """Return historical generated articles/copies.""" + query = ( + select(Copy) + .order_by(desc(Copy.created_at)) + .offset(offset) + .limit(limit) + ) + result = await db.execute(query) + items = result.scalars().all() + + return { + "items": [ + { + "id": str(item.id), + "bundle_id": str(item.bundle_id), + "campaign_id": str(item.campaign_id), + "variant_label": item.variant_label, + "title": item.hook, + "body": item.body, + "channel": item.channel, + "status": item.status, + "created_at": item.created_at.isoformat() + } + for item in items + ] + } + + +@router.delete("/{article_id}", summary="Delete an article by ID") +async def delete_article(article_id: UUID, db: AsyncSession = Depends(get_db)): + result = await db.execute(select(Copy).where(Copy.id == article_id)) + item = result.scalar_one_or_none() + if item is None: + raise HTTPException(status_code=404, detail="Article not found") + await db.execute(delete(Copy).where(Copy.id == article_id)) + await db.commit() + return {"deleted": str(article_id)} diff --git a/backend/app/api/router.py b/backend/app/api/router.py index 29171e6..c89057c 100644 --- a/backend/app/api/router.py +++ b/backend/app/api/router.py @@ -3,10 +3,12 @@ from .campaigns import router as campaigns_router from .agents import router as agents_router +from .articles import router as articles_router from .ws import router as ws_router api_router = APIRouter() api_router.include_router(campaigns_router, prefix="/v1/campaigns", tags=["Campaigns"]) api_router.include_router(agents_router, prefix="/v1/agents", tags=["A2A Agents"]) +api_router.include_router(articles_router, prefix="/v1/articles", tags=["Articles"]) api_router.include_router(ws_router, tags=["WebSocket"]) diff --git a/backend/app/config.py b/backend/app/config.py index 752078a..45f0ea5 100644 --- a/backend/app/config.py +++ b/backend/app/config.py @@ -70,6 +70,10 @@ class Settings(BaseSettings): google_ads_client_secret: str = Field(default="", alias="GOOGLE_ADS_CLIENT_SECRET") google_ads_refresh_token: str = Field(default="", alias="GOOGLE_ADS_REFRESH_TOKEN") + # ── Zhihu ────────────────────────────────────────────────────── + zhihu_cookie: str = Field(default="", alias="ZHIHU_COOKIE") + zhihu_zst_81: str = Field(default="", alias="ZHIHU_ZST_81") + # ── Image Generation ─────────────────────────────────────────── openai_api_key: str = Field(default="", alias="OPENAI_API_KEY") stability_api_key: str = Field(default="", alias="STABILITY_API_KEY") diff --git a/backend/app/core/llm.py b/backend/app/core/llm.py index 84fd5cb..4133765 100644 --- a/backend/app/core/llm.py +++ b/backend/app/core/llm.py @@ -81,16 +81,23 @@ async def chat_completion( raise ValueError(f"Unsupported provider: {provider}") async def _anthropic_completion(self, messages, system, model, max_tokens): + resolved_model = model or settings.anthropic_model + logger.info("llm_request", provider="anthropic", model=resolved_model, + system=system, messages=messages) response = await self.anthropic.messages.create( - model=model or settings.anthropic_model, + model=resolved_model, max_tokens=max_tokens or settings.anthropic_max_tokens, system=system, messages=messages, ) - return response.content[0].text + result = response.content[0].text + logger.info("llm_response", provider="anthropic", model=resolved_model, + response=result) + return result async def _openai_compatible_completion(self, base_url, api_key, messages, system, model, max_tokens): - async with httpx.AsyncClient() as client: + # NOTE: Timeout must be 180s for long technical articles + async with httpx.AsyncClient(timeout=180.0) as client: full_messages = [] if system: full_messages.append({"role": "system", "content": system}) @@ -113,9 +120,14 @@ async def _openai_compatible_completion(self, base_url, api_key, messages, syste else: url = url.rstrip("/") + "/chat/completions" - response = await client.post(url, json=payload, headers=headers, timeout=60.0) + logger.info("llm_request", provider=url.split("/")[2], model=model, + messages=full_messages) + response = await client.post(url, json=payload, headers=headers) response.raise_for_status() data = response.json() - return data["choices"][0]["message"]["content"] + result = data["choices"][0]["message"]["content"] + logger.info("llm_response", provider=url.split("/")[2], model=model, + response=result) + return result llm_client = LLMClient() diff --git a/backend/app/tasks/agent_tasks.py b/backend/app/tasks/agent_tasks.py index c971d5f..81630ce 100644 --- a/backend/app/tasks/agent_tasks.py +++ b/backend/app/tasks/agent_tasks.py @@ -10,10 +10,25 @@ from arq.connections import RedisSettings from app.config import settings +from app.core.event_bus import event_bus logger = structlog.get_logger(__name__) +# ── Lifecycle Hooks ────────────────────────────────────────────────────────── + +async def startup(ctx: dict): + """Initialize resources for the worker process.""" + await event_bus.connect() + logger.info("worker_startup_complete") + + +async def shutdown(ctx: dict): + """Cleanup resources.""" + await event_bus.disconnect() + logger.info("worker_shutdown_complete") + + # ── Task Functions (executed by ARQ worker) ─────────────────────────────────── async def run_campaign_pipeline(ctx: dict, campaign_id: str): @@ -21,6 +36,10 @@ async def run_campaign_pipeline(ctx: dict, campaign_id: str): Full campaign pipeline: PLANNING → DEPLOYED → MONITORING → OPTIMIZING. Invokes the LangGraph StateGraph with PostgreSQL checkpointer. """ + # Robustness: ensure event_bus is connected in this worker process + if not event_bus._redis: + await event_bus.connect() + logger.info("campaign_pipeline_start", campaign_id=campaign_id) from app.database import get_checkpointer, async_session_factory @@ -50,7 +69,14 @@ async def run_campaign_pipeline(ctx: dict, campaign_id: str): config = {"configurable": {"thread_id": campaign_id}} result = await graph.ainvoke(initial_state, config=config) - logger.info("campaign_pipeline_done", campaign_id=campaign_id, status=result.get("status")) + # Final status update to trigger frontend 'COMPLETED' (loop-back) visual + final_status = result.get("status", "COMPLETED") + if final_status == "OPTIMIZING": + # In the logic, OPTIMIZING means we finished a loop and KPI was met + # Let's broadcast COMPLETED to trigger the UI loop-back animation + await event_bus.publish("StatusChanged", {"old_status": "OPTIMIZING", "new_status": "COMPLETED"}, campaign_id) + + logger.info("campaign_pipeline_done", campaign_id=campaign_id, status=final_status) return result @@ -142,6 +168,8 @@ async def cancel_job(task_id: str) -> bool: class WorkerSettings: functions = [run_campaign_pipeline, run_agent_node] + on_startup = startup + on_shutdown = shutdown redis_settings = RedisSettings.from_dsn(settings.arq_redis_url) max_jobs = settings.arq_max_jobs job_timeout = settings.arq_job_timeout diff --git a/backend/openautogrowth_backend.egg-info/PKG-INFO b/backend/openautogrowth_backend.egg-info/PKG-INFO new file mode 100644 index 0000000..d23cde0 --- /dev/null +++ b/backend/openautogrowth_backend.egg-info/PKG-INFO @@ -0,0 +1,32 @@ +Metadata-Version: 2.4 +Name: openautogrowth-backend +Version: 1.0.0 +Summary: OpenAutoGrowth — AI Multi-Agent Growth Engine Backend +Requires-Python: >=3.12 +Requires-Dist: fastapi<0.116,>=0.115 +Requires-Dist: uvicorn[standard]<0.33,>=0.32 +Requires-Dist: pydantic<3.0,>=2.9 +Requires-Dist: pydantic-settings<3.0,>=2.6 +Requires-Dist: sqlalchemy[asyncio]<3.0,>=2.0 +Requires-Dist: asyncpg<0.31,>=0.30 +Requires-Dist: alembic<2.0,>=1.14 +Requires-Dist: langgraph>=0.2 +Requires-Dist: langchain-anthropic>=0.3 +Requires-Dist: langgraph-checkpoint-postgres>=2.0 +Requires-Dist: anthropic<0.50,>=0.40 +Requires-Dist: httpx<0.29,>=0.28 +Requires-Dist: mcp<2.0,>=1.0 +Requires-Dist: redis[hiredis]<6.0,>=5.2 +Requires-Dist: arq<0.27,>=0.26 +Requires-Dist: python-dotenv<2.0,>=1.0 +Requires-Dist: structlog<25.0,>=24.4 +Requires-Dist: tenacity<10.0,>=9.0 +Requires-Dist: python-jose[cryptography]<4.0,>=3.3 +Requires-Dist: passlib[bcrypt]<2.0,>=1.7 +Provides-Extra: dev +Requires-Dist: pytest>=8.0; extra == "dev" +Requires-Dist: pytest-asyncio>=0.24; extra == "dev" +Requires-Dist: pytest-cov>=6.0; extra == "dev" +Requires-Dist: httpx>=0.28; extra == "dev" +Requires-Dist: ruff>=0.8; extra == "dev" +Requires-Dist: mypy>=1.13; extra == "dev" diff --git a/backend/openautogrowth_backend.egg-info/SOURCES.txt b/backend/openautogrowth_backend.egg-info/SOURCES.txt new file mode 100644 index 0000000..c351f90 --- /dev/null +++ b/backend/openautogrowth_backend.egg-info/SOURCES.txt @@ -0,0 +1,46 @@ +pyproject.toml +app/__init__.py +app/config.py +app/database.py +app/agents/__init__.py +app/agents/analysis.py +app/agents/channel_exec.py +app/agents/content_gen.py +app/agents/graph.py +app/agents/multimodal.py +app/agents/optimizer.py +app/agents/planner.py +app/agents/state.py +app/agents/strategy.py +app/api/__init__.py +app/api/agents.py +app/api/campaigns.py +app/api/router.py +app/api/ws.py +app/core/__init__.py +app/core/event_bus.py +app/core/llm.py +app/core/memory.py +app/core/rule_engine.py +app/models/__init__.py +app/models/analytics.py +app/models/campaign.py +app/models/content.py +app/models/optimization.py +app/models/user.py +app/protocols/__init__.py +app/protocols/a2a/__init__.py +app/protocols/a2a/models.py +app/protocols/mcp/__init__.py +app/protocols/mcp/tools.py +app/schemas/__init__.py +app/schemas/a2a.py +app/schemas/agent.py +app/schemas/campaign.py +app/tasks/__init__.py +app/tasks/agent_tasks.py +openautogrowth_backend.egg-info/PKG-INFO +openautogrowth_backend.egg-info/SOURCES.txt +openautogrowth_backend.egg-info/dependency_links.txt +openautogrowth_backend.egg-info/requires.txt +openautogrowth_backend.egg-info/top_level.txt \ No newline at end of file diff --git a/backend/openautogrowth_backend.egg-info/dependency_links.txt b/backend/openautogrowth_backend.egg-info/dependency_links.txt new file mode 100644 index 0000000..8b13789 --- /dev/null +++ b/backend/openautogrowth_backend.egg-info/dependency_links.txt @@ -0,0 +1 @@ + diff --git a/backend/openautogrowth_backend.egg-info/requires.txt b/backend/openautogrowth_backend.egg-info/requires.txt new file mode 100644 index 0000000..16d1ba6 --- /dev/null +++ b/backend/openautogrowth_backend.egg-info/requires.txt @@ -0,0 +1,28 @@ +fastapi<0.116,>=0.115 +uvicorn[standard]<0.33,>=0.32 +pydantic<3.0,>=2.9 +pydantic-settings<3.0,>=2.6 +sqlalchemy[asyncio]<3.0,>=2.0 +asyncpg<0.31,>=0.30 +alembic<2.0,>=1.14 +langgraph>=0.2 +langchain-anthropic>=0.3 +langgraph-checkpoint-postgres>=2.0 +anthropic<0.50,>=0.40 +httpx<0.29,>=0.28 +mcp<2.0,>=1.0 +redis[hiredis]<6.0,>=5.2 +arq<0.27,>=0.26 +python-dotenv<2.0,>=1.0 +structlog<25.0,>=24.4 +tenacity<10.0,>=9.0 +python-jose[cryptography]<4.0,>=3.3 +passlib[bcrypt]<2.0,>=1.7 + +[dev] +pytest>=8.0 +pytest-asyncio>=0.24 +pytest-cov>=6.0 +httpx>=0.28 +ruff>=0.8 +mypy>=1.13 diff --git a/backend/openautogrowth_backend.egg-info/top_level.txt b/backend/openautogrowth_backend.egg-info/top_level.txt new file mode 100644 index 0000000..b80f0bd --- /dev/null +++ b/backend/openautogrowth_backend.egg-info/top_level.txt @@ -0,0 +1 @@ +app diff --git a/backend/pyproject.toml b/backend/pyproject.toml index 70df77c..c63c097 100644 --- a/backend/pyproject.toml +++ b/backend/pyproject.toml @@ -23,8 +23,9 @@ dependencies = [ "alembic>=1.14,<2.0", # Agent framework - "langgraph>=0.2,<0.3", - "langchain-anthropic>=0.3,<0.4", + "langgraph>=0.2", + "langchain-anthropic>=0.3", + "langgraph-checkpoint-postgres>=2.0", "anthropic>=0.40,<0.50", # Protocol diff --git a/index.html b/index.html index 1ac3b1f..8c16a13 100644 --- a/index.html +++ b/index.html @@ -20,7 +20,7 @@ OpenAutoGrowth
- Multi-Agent · Closed-Loop · v2.0 + Multi-Agent · Closed-Loop · v0.0.1
@@ -66,6 +66,7 @@

Content Generation

+
@@ -158,44 +159,44 @@

🔍 System Activity Log

Agent Pipeline

-
+
🧠 Orchestrator
-
→
-
+
→
+
🗺️ Planner
-
→
-
+
→
+
+ 🎯 + Strategy +
+
📝 ContentGen
-
+
🎨 Multimodal
-
- 🎯 - Strategy -
-
→
-
+
→
+
📡 ChannelExec
-
→
-
+
→
+
📊 Analysis
-
→
-
+
→
+
⚙️ Optimizer
-
↩ Loop
+
↩ Loop
@@ -205,6 +206,87 @@

Agent Pipeline

+ + + + + + + +
+
+
+

📝 Preview Article

+ +
+
+
+ + +
+
+ + +
+
+
+ + +
+
+
+ diff --git a/main.js b/main.js index ec4912c..7aeb548 100644 --- a/main.js +++ b/main.js @@ -69,6 +69,18 @@ class DashboardController { constructor() { this.activeCampaignId = null; this.logLines = 0; + this.launchType = 'campaign'; // 'campaign' or 'article' + this.pipelineMapping = { + 'PLANNING': ['node-orchestrator', 'node-planner', 'arrow-1', 'arrow-2'], + 'STRATEGY': ['node-strategy', 'arrow-2'], + 'CONTENT_GEN': ['node-contentgen', 'arrow-2'], + 'MULTIMODAL': ['node-multimodal', 'arrow-2'], + 'EXECUTING': ['node-channelexec', 'arrow-3'], + 'DEPLOYED': ['node-channelexec', 'arrow-3'], + 'ANALYZING': ['node-analysis', 'arrow-4'], + 'OPTIMIZING': ['node-optimizer', 'arrow-5'], + 'COMPLETED': ['arrow-loop'] + }; } init() { @@ -77,6 +89,36 @@ class DashboardController { this._subscribeToEvents(); this._startAnimations(); this.log(i18n.t('status_online'), 'success'); + this._updatePipeline('IDLE'); // Initial state should be idle + } + + _updatePipeline(status) { + console.log(`[UI] Updating pipeline for status: ${status}`); + + // Define which nodes should be ON for each stage (Cumulative) + const stageProgress = { + 'IDLE': [], + 'PLANNING': ['node-orchestrator', 'node-planner', 'arrow-1'], + 'STRATEGY': ['node-orchestrator', 'node-planner', 'node-strategy', 'arrow-1', 'arrow-2'], + 'CONTENT_GEN': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'arrow-1', 'arrow-2'], + 'MULTIMODAL': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'arrow-1', 'arrow-2'], + 'DEPLOYED': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'node-channelexec', 'arrow-1', 'arrow-2', 'arrow-3'], + 'EXECUTING': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'node-channelexec', 'arrow-1', 'arrow-2', 'arrow-3'], + 'ANALYZING': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'node-channelexec', 'node-analysis', 'arrow-1', 'arrow-2', 'arrow-3', 'arrow-4'], + 'OPTIMIZING': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'node-channelexec', 'node-analysis', 'node-optimizer', 'arrow-1', 'arrow-2', 'arrow-3', 'arrow-4', 'arrow-5'], + 'COMPLETED': ['node-orchestrator', 'node-planner', 'node-strategy', 'node-contentgen', 'node-multimodal', 'node-channelexec', 'node-analysis', 'node-optimizer', 'arrow-1', 'arrow-2', 'arrow-3', 'arrow-4', 'arrow-5', 'arrow-loop'] + }; + + const activeIds = stageProgress[status] || []; + + // If it's IDLE, clear all. Otherwise, we only add. + if (status === 'IDLE') { + document.querySelectorAll('.pipeline-node, .pipeline-arrow').forEach(el => el.classList.remove('active')); + } + + activeIds.forEach(id => { + document.getElementById(id)?.classList.add('active'); + }); } // ── Event subscriptions → UI updates ──────────────────────── @@ -91,6 +133,7 @@ class DashboardController { globalEventBus.subscribe('StatusChanged', ({ payload: { old_status, new_status }, campaign_id }) => { this.log(i18n.t('log_campaign_status_change', { id: campaign_id, old: old_status, new: new_status }), 'info'); this._updateCampaignBadge(new_status); + this._updatePipeline(new_status); }); globalEventBus.subscribe('PlanGenerated', ({ payload: { plan } }) => { @@ -146,10 +189,25 @@ class DashboardController { // Card action buttons document.getElementById('btn-gen-new')?.addEventListener('click', () => this._triggerContentGen()); + document.getElementById('btn-view-history')?.addEventListener('click', () => this._showHistoryModal()); document.getElementById('btn-exec')?.addEventListener('click', () => this._triggerExecution()); document.getElementById('btn-sync')?.addEventListener('click', () => this._triggerAnalysis()); document.getElementById('btn-optimize')?.addEventListener('click', () => this._triggerOptimizer()); + // Article Modal actions + document.getElementById('btn-close-modal')?.addEventListener('click', () => this._closeArticleModal()); + document.getElementById('btn-cancel-article')?.addEventListener('click', () => this._closeArticleModal()); + document.getElementById('btn-publish-article')?.addEventListener('click', () => this._publishArticle()); + + // Launch Modal actions + document.getElementById('btn-close-launch-modal')?.addEventListener('click', () => this._closeLaunchModal()); + document.getElementById('btn-cancel-launch')?.addEventListener('click', () => this._closeLaunchModal()); + document.getElementById('btn-confirm-launch')?.addEventListener('click', () => this._confirmLaunchPromotion()); + + // History Modal actions + document.getElementById('btn-close-history-modal')?.addEventListener('click', () => this._closeHistoryModal()); + document.getElementById('btn-close-history')?.addEventListener('click', () => this._closeHistoryModal()); + // Language buttons document.getElementById('btn-lang-zh')?.addEventListener('click', () => i18n.setLocale('zh')); document.getElementById('btn-lang-en')?.addEventListener('click', () => i18n.setLocale('en')); @@ -158,20 +216,139 @@ class DashboardController { // ── Actions ────────────────────────────────────────────────── async _launchCampaign() { + this.launchType = 'campaign'; + document.getElementById('launch-modal').style.display = 'block'; + document.getElementById('promo-goal-input').value = '新品 X Pro 冷启动推广，Q2 GMV 达 500 万'; + document.getElementById('btn-confirm-launch').textContent = '🚀 Launch Full Pipeline'; + } + + async _triggerContentGen() { + this.launchType = 'article'; + this._updatePipeline('CONTENT_GEN'); + document.getElementById('launch-modal').style.display = 'block'; + document.getElementById('promo-goal-input').value = 'Promote my open source project https://github.com/CadanHu/data-analyse-system on Zhihu'; + document.getElementById('btn-confirm-launch').textContent = '🚀 Start Generation'; + } + + _closeLaunchModal() { + document.getElementById('launch-modal').style.display = 'none'; + } + + async _showHistoryModal() { + const historyList = document.getElementById('history-list'); + historyList.innerHTML = '
Loading history...
'; + document.getElementById('history-modal').style.display = 'block'; + + const response = await api.listArticles(); + if (response.success) { + this._renderHistoryList(response.data.items); + } else { + historyList.innerHTML = `
Failed to load history: ${response.error}
`; + } + } + + _renderHistoryList(items) { + const historyList = document.getElementById('history-list'); + if (!items || items.length === 0) { + historyList.innerHTML = '
No history found.
'; + return; + } + + historyList.innerHTML = ''; + items.forEach(item => { + const div = document.createElement('div'); + div.className = 'history-item'; + div.innerHTML = ` +

${item.title || 'Untitled'}

+
+ ${new Date(item.created_at).toLocaleDateString()} + ${item.status} + +
+ `; + div.querySelector('h4').onclick = () => { + this._closeHistoryModal(); + this._showArticleModal(item); + }; + div.querySelector('.history-delete-btn').onclick = async (e) => { + e.stopPropagation(); + if (!confirm(`Delete "${item.title || 'Untitled'}"?`)) return; + const res = await api.deleteArticle(item.id); + if (res.success) { + div.remove(); + if (historyList.children.length === 0) { + historyList.innerHTML = '
No history found.
'; + } + } else { + alert('Delete failed: ' + res.error); + } + }; + historyList.appendChild(div); + }); + } + + _closeHistoryModal() { + document.getElementById('history-modal').style.display = 'none'; + } + + async _confirmLaunchPromotion() { + const goal = document.getElementById('promo-goal-input').value; + const budget = parseInt(document.getElementById('promo-budget-input')?.value || '50000'); + const kpi = parseFloat(document.getElementById('promo-kpi-input')?.value || '3.0'); + const selectedChannels = Array.from(document.querySelectorAll('input[name="channel"]:checked')).map(cb => cb.value); + + if (!goal || selectedChannels.length === 0) { + alert('Please provide a goal and at least one channel.'); + return; + } + + this._closeLaunchModal(); + + if (this.launchType === 'article') { + await this._runArticleGenWorkflow(goal, selectedChannels); + } else { + await this._runFullPipelineWorkflow(goal, budget, kpi, selectedChannels); + } + } + + async _runArticleGenWorkflow(goal, channels) { + this._setButtonState('btn-gen-new', i18n.t('btn_generating'), true); + this.log('Starting AI technical article generation...', 'info'); + + const response = await api.callAgent('content_gen', { + campaign_id: this.activeCampaignId || 'demo', + goal: goal, + strategy: { channel_plan: channels.map(ch => ({ channel: ch })) }, + kpi: { metric: 'awareness', target: 'high' } + }); + + if (response.success) { + const output = response.data; + if (output.content?.variants?.[0]?.body) { + this._showArticleModal(output.content.variants[0]); + } + } else { + this.log(`Content generation failed: ${response.error}`, 'error'); + } + this._setButtonState('btn-gen-new', i18n.t('btn_gen_new'), false); + } + + async _runFullPipelineWorkflow(goal, budget, kpi, channels) { this._setButtonState('btn-launch', i18n.t('btn_launching'), true); this.log(i18n.t('log_campaign_cycle_started'), 'divider'); + this._updatePipeline('PLANNING'); const response = await api.createCampaign({ - goal: '新品 X Pro 冷启动推广，Q2 GMV 达 500 万', - budget: { total: 50000, currency: 'CNY', daily_cap: 5000 }, - timeline: { start: '2026-04-10', end: '2026-04-30' }, - kpi: { metric: 'ROAS', target: 3.0 }, - constraints: { channels: ['tiktok', 'meta', 'google'], region: 'CN' }, + goal: goal, + budget: { total: budget, currency: 'CNY', daily_cap: Math.floor(budget / 10) }, + timeline: { start: new Date().toISOString().split('T')[0], end: '2026-12-31' }, + kpi: { metric: 'ROAS', target: kpi }, + constraints: { channels: channels, region: 'CN' }, }); if (!response.success) { this.log(`Error: ${response.error}`, 'error'); - this._setButtonState('btn-launch', 'Launch Campaign', false); + this._setButtonState('btn-launch', i18n.t('btn_launch'), false); return; } @@ -179,13 +356,13 @@ class DashboardController { this.log(i18n.t('log_campaign_created', { id: this.activeCampaignId.slice(0, 8) }), 'success'); this._updateCampaignBadge(response.data.status); - // 订阅后端 WebSocket 推送 → 驱动 UI + // Subscribe to WS wsBroadcaster.subscribe(this.activeCampaignId, (msg) => { console.log('[WS →]', msg.type, msg.payload); this._handleWsMessage(msg); }); - // 触发规划启动 + // Trigger Start const startResp = await api.startCampaign(this.activeCampaignId); if (startResp.success) { this._updateCampaignBadge('PLANNING'); @@ -196,21 +373,42 @@ class DashboardController { this._setButtonState('btn-launch', i18n.t('btn_launch'), false); } - async _triggerContentGen() { - this._setButtonState('btn-gen-new', i18n.t('btn_generating'), true); - const agent = new ContentGenAgent(); - await agent.run({ - product: { name: 'X Pro', category: 'SaaS', USP: ['AI驱动', '一键部署', '成本降低60%'] }, - target_persona: { age: '25-35', interest: ['创业', '效率工具'] }, - channels: ['tiktok', 'weibo'], - tone: 'energetic', - ab_variants: 3, - campaign_id: this.activeCampaignId || 'demo', + _showArticleModal(variant) { + document.getElementById('article-title-input').value = variant.title || 'Untitled Article'; + document.getElementById('article-body-input').value = variant.body || ''; + document.getElementById('article-modal').style.display = 'block'; + } + + _closeArticleModal() { + document.getElementById('article-modal').style.display = 'none'; + } + + async _publishArticle() { + this._closeArticleModal(); + this.log('Publishing article to Zhihu...', 'info'); + + const title = document.getElementById('article-title-input').value; + const body = document.getElementById('article-body-input').value; + + // Trigger execution agent for Zhihu via Python backend + // We don't await polling here to keep UI responsive as requested + api.callAgent('channel_exec', { + campaign_id: this.activeCampaignId || 'demo', + strategy: { channel_plan: [{ channel: 'zhihu' }] }, + content: { variants: [{ title, body, channel: 'zhihu' }] } + }).then(response => { + if (response.success) { + this.log('Article successfully published to Zhihu!', 'success'); + } else { + this.log(`Publishing failed: ${response.error}`, 'error'); + } }); - this._setButtonState('btn-gen-new', i18n.t('btn_gen_new'), false); + + this.log('Article queued for publishing to Zhihu', 'success'); } async _triggerExecution() { + this._updatePipeline('EXECUTING'); this._setButtonState('btn-exec', i18n.t('btn_executing'), true); const agent = new ChannelExecAgent(); await agent.run({ @@ -225,6 +423,7 @@ class DashboardController { } async _triggerAnalysis() { + this._updatePipeline('ANALYZING'); this._setButtonState('btn-sync', i18n.t('btn_syncing'), true); const agent = new AnalysisAgent(); await agent.run({ metrics: ['CTR', 'ROAS'], campaign_id: this.activeCampaignId || 'demo' }); @@ -232,6 +431,7 @@ class DashboardController { } async _triggerOptimizer() { + this._updatePipeline('OPTIMIZING'); this._setButtonState('btn-optimize', i18n.t('btn_optimizing'), true); const analysisAgent = new AnalysisAgent(); const report = await analysisAgent.run({ metrics: ['CTR', 'ROAS'], campaign_id: this.activeCampaignId || 'demo' }); @@ -248,36 +448,44 @@ class DashboardController { case 'campaign.status_changed': this.log(i18n.t('log_campaign_status_change', { id: campaign_id?.slice(0,8), old: payload.old_status, new: payload.new_status }), 'info'); this._updateCampaignBadge(payload.new_status); + this._updatePipeline(payload.new_status); break; case 'campaign.plan_ready': this.log(i18n.t('log_planner_dag', { n: payload.plan?.tasks?.length ?? '?', s: payload.plan?.scenario }), 'info'); this._updateStat('gen-count', `${payload.plan?.tasks?.length ?? '?'} Tasks`); + this._updatePipeline('PLANNING'); break; case 'task.content_generated': this.log(i18n.t('log_content_gen', { n: payload.bundle?.variants?.length ?? 0 }), 'success'); this._updateStat('gen-count', `${payload.bundle?.variants?.length ?? 0} Variants`); + this._updatePipeline('CONTENT_GEN'); break; case 'task.assets_generated': this.log(i18n.t('log_assets_gen', { n: payload.asset_ids?.length ?? 0, type: payload.type }), 'success'); + this._updatePipeline('MULTIMODAL'); break; case 'task.strategy_decided': this.log(i18n.t('log_strategy_decided', { channels: payload.strategy?.channel_plan?.map(c => c.channel).join(', ') }), 'info'); + this._updatePipeline('STRATEGY'); break; case 'task.ad_deployed': const platforms = payload.platforms?.join(', '); this.log(i18n.t('log_ads_deployed', { platforms }), 'success'); this._addFeedItem(i18n.t('log_deployed_to', { platforms: payload.platforms?.join(' & ') })); this._updateStat('exec-reach', '87.4%'); + this._updatePipeline('DEPLOYED'); break; case 'metrics.updated': this.log(i18n.t('log_analytics_report', { roas: payload.metrics?.roas?.toFixed(2), ctr: ((payload.metrics?.ctr ?? 0) * 100).toFixed(2) }), 'info'); this._updateStat('roi-value', `+${((payload.metrics?.roas - 1) * 100)?.toFixed(1)}%`); this._animateChartBars(payload.metrics?.roas); + this._updatePipeline('ANALYZING'); break; case 'optimization.applied': const types = payload.actions?.map(a => a.type).join(', ') || 'NONE'; this.log(i18n.t('log_optimizer_fired', { types }), 'warning'); this._updateOptStatus(i18n.t('log_opt_status', { loop: payload.loop_count, types })); + this._updatePipeline('OPTIMIZING'); break; case 'anomaly.detected': this.log(i18n.t('log_anomaly_detected', { metric: payload.metric, channel: payload.channel, severity: payload.severity }), 'error'); diff --git a/src/agents/ContentGen.js b/src/agents/ContentGen.js index 62a1ddc..4be195a 100644 --- a/src/agents/ContentGen.js +++ b/src/agents/ContentGen.js @@ -41,10 +41,23 @@ export class ContentGenAgent { return output; } - async _generateVariants({ product, target_persona, tone, ab_variants }) { + async _generateVariants({ product, target_persona, channels, tone, ab_variants }) { // 模拟 LLM 生成（生产环境替换为真实 API 调用） await this._simulateLatency(1500); + const isTechnicalPromo = channels?.includes('zhihu') || channels?.includes('juejin'); + + if (isTechnicalPromo) { + return [{ + id: 'copy_vA', + variant: 'A', + title: `深入浅出 ${product?.name || '开源项目'}：重构企业级数据分析工作流`, + body: `# ${product?.name || 'DataPulse'} 技术解析\n\n在现代数据驱动的商业环境中，我们需要更智能的工具。${product?.name} 正是为此而生。\n\n## 核心特性\n- ${product?.USP?.join('\n- ')}\n\n## 为什么选择 ${product?.name}？\n因为它不仅是一个工具，更是一个完整的生态。`, + channel: 'zhihu', + llm_model: 'claude-3-5-sonnet-mock', + }]; + } + const hooks = [ `${product?.USP?.[0] || '全新功能'}，让你效率翻倍`, `${target_persona?.age || '年轻人'}都在用的${product?.name || '好工具'}`, diff --git a/src/api/routes.js b/src/api/routes.js index acaa5cd..3d960a7 100644 --- a/src/api/routes.js +++ b/src/api/routes.js @@ -60,6 +60,67 @@ export class CampaignAPI { return this._request('GET', `/campaigns/${id}/events`); } + // ── Articles ────────────────────────────────────────────────────────── + + async listArticles(limit = 20, offset = 0) { + return this._request('GET', `/articles?limit=${limit}&offset=${offset}`); + } + + async deleteArticle(id) { + return this._request('DELETE', `/articles/${id}`); + } + + // ── A2A Agents ──────────────────────────────────────────────────────── + + /** + * Call a backend agent directly (A2A style) + */ + async callAgent(agentName, input) { + const taskId = `task_${Math.random().toString(36).slice(2, 10)}`; + const payload = { + id: taskId, + message: { + role: 'user', + parts: [{ type: 'text', text: JSON.stringify(input) }] + } + }; + + const submitResp = await this._request('POST', `/agents/${agentName}/tasks/send`, payload); + if (!submitResp.success) return submitResp; + + // Simple polling for result + return this._pollTask(agentName, taskId); + } + + async _pollTask(agentName, taskId, retry = 0) { + // Timeout: 150 * 2s = 300 seconds + if (retry > 150) return { success: false, error: 'Polling timeout' }; + + await new Promise(r => setTimeout(r, 2000)); + const resp = await this._request('GET', `/agents/${agentName}/tasks/${taskId}`); + + if (!resp.success) return resp; + + const state = resp.data.status?.state; + if (state === 'completed') { + const artifact = resp.data.artifacts?.find(a => a.name === 'result'); + const text = artifact?.parts?.find(p => p.type === 'text')?.text; + const data = text ? JSON.parse(text) : {}; + + // Handle internal agent errors + if (data.errors && data.errors.length > 0) { + return { success: false, error: data.errors[0].error || 'Agent execution failed' }; + } + + return { success: true, data: data }; + } else if (state === 'failed' || state === 'canceled') { + const error = resp.data.metadata?.error || `Task ${state}`; + return { success: false, error: error }; + } + + return this._pollTask(agentName, taskId, retry + 1); + } + // ── Internal fetch helper ───────────────────────────────────────────── async _request(method, path, body = null) { diff --git a/src/i18n/locales/en.js b/src/i18n/locales/en.js index 5bb4a77..b922570 100644 --- a/src/i18n/locales/en.js +++ b/src/i18n/locales/en.js @@ -1,6 +1,6 @@ export default { "app_title": "OpenAutoGrowth | AI Growth Closed-Loop Engine", - "nav_version": "Multi-Agent · Closed-Loop · v2.0", + "nav_version": "Multi-Agent · Closed-Loop · v0.0.1", "nav_no_campaign": "NO CAMPAIGN", "status_no campaign": "NO CAMPAIGN", "status_planning": "PLANNING", diff --git a/src/i18n/locales/zh.js b/src/i18n/locales/zh.js index 85c2e0c..c24e43c 100644 --- a/src/i18n/locales/zh.js +++ b/src/i18n/locales/zh.js @@ -1,6 +1,6 @@ export default { "app_title": "OpenAutoGrowth | AI 增长闭环引擎", - "nav_version": "多智能体 · 闭环 · v2.0", + "nav_version": "多智能体 · 闭环 · v0.0.1", "nav_no_campaign": "无活动", "status_no campaign": "无活动", "status_planning": "规划中", @@ -17,6 +17,7 @@ export default { "card_gen_label": "产出资源", "card_gen_desc": "ContentGen + 多模态智能体并行生成 A/B 测试文案和视觉资源。", "btn_view_library": "查看库", + "btn_history": "历史记录", "btn_gen_new": "生成新内容", "card_exec_title": "智能投放", @@ -63,5 +64,15 @@ export default { "log_optimizer_fired": "优化器已触发：{types}", "log_anomaly_detected": "⚠️ 异常：在 {channel} 上的 {metric} ({severity})", "log_opt_status": "循环 #{loop} — {types}", - "log_campaign_status_change": "活动 {id}: {old} → {new}" + "log_campaign_status_change": "活动 {id}: {old} → {new}", + "modal_article_title": "预览文章", + "label_article_title": "标题", + "label_article_body": "正文 (Markdown)", + "btn_cancel": "取消", + "btn_publish": "🚀 发布到知乎", + "modal_launch_title": "发起新推广", + "label_promo_goal": "推广目标 (包含 GitHub URL)", + "btn_confirm_launch": "🚀 开始生成", + "modal_history_title": "文章历史记录", + "btn_close": "关闭" }; diff --git a/style.css b/style.css index dea1fb8..4bbee95 100644 --- a/style.css +++ b/style.css @@ -414,32 +414,49 @@ html[lang="en"] #btn-lang-en { flex-direction: column; align-items: center; gap: 0.25rem; - background: rgba(99,102,241,0.1); - border: 1px solid rgba(99,102,241,0.25); + background: rgba(255, 255, 255, 0.03); + border: 1px solid rgba(255, 255, 255, 0.1); border-radius: 10px; padding: 0.5rem 0.75rem; min-width: 80px; - transition: var(--transition); + transition: all 0.5s cubic-bezier(0.4, 0, 0.2, 1); + opacity: 0.5; + filter: grayscale(1); } -.pipeline-node:hover { - background: rgba(99,102,241,0.2); - border-color: rgba(99,102,241,0.5); - transform: translateY(-2px); +.pipeline-node.active { + background: rgba(99, 102, 241, 0.15); + border-color: rgba(99, 102, 241, 0.5); + opacity: 1; + filter: grayscale(0); + transform: translateY(-4px); + box-shadow: 0 10px 20px rgba(99, 102, 241, 0.2); } -.pipeline-node.parallel { - background: rgba(16,185,129,0.08); - border-color: rgba(16,185,129,0.25); +.pipeline-node.active.parallel { + background: rgba(16, 185, 129, 0.15); + border-color: rgba(16, 185, 129, 0.5); + box-shadow: 0 10px 20px rgba(16, 185, 129, 0.2); } -.p-icon { font-size: 1.2rem; } -.p-label { font-size: 0.7rem; color: var(--text-secondary); text-align: center; white-space: nowrap; } - .pipeline-arrow { color: var(--text-secondary); font-size: 1.2rem; flex-shrink: 0; + transition: all 0.3s; + opacity: 0.3; +} + +.pipeline-arrow.active { + color: var(--text-primary); + opacity: 1; + animation: arrow-pulse 1.5s infinite; +} + +@keyframes arrow-pulse { + 0% { transform: translateX(0); opacity: 0.5; } + 50% { transform: translateX(5px); opacity: 1; text-shadow: 0 0 8px white; } + 100% { transform: translateX(0); opacity: 0.5; } } .pipeline-arrow.loop-back { @@ -461,6 +478,166 @@ html[lang="en"] #btn-lang-en { font-size: 0.8rem; } +/* ── Modals ────────────────────────────────────────────────── */ +.modal { + display: none; + position: fixed; + z-index: 1000; + left: 0; top: 0; + width: 100%; height: 100%; + background-color: rgba(0,0,0,0.85); + backdrop-filter: blur(10px); +} + +.modal-content { + background: var(--bg-color); + margin: 5% auto; + padding: 2rem; + border: 1px solid var(--border-color); + width: 80%; + max-width: 900px; + border-radius: var(--radius-lg); + box-shadow: 0 25px 50px rgba(0,0,0,0.5); +} + +.modal-header { + display: flex; + justify-content: space-between; + align-items: center; + margin-bottom: 1.5rem; +} + +.close-btn { + background: none; + border: none; + color: var(--text-secondary); + font-size: 2rem; + cursor: pointer; +} + +.modal-body { + display: flex; + flex-direction: column; + gap: 1.5rem; +} + +.input-group { + display: flex; + flex-direction: column; + gap: 0.5rem; +} + +.checkbox-group { + display: flex; + gap: 1.5rem; + padding: 0.5rem 0; +} + +.checkbox-group label { + display: flex; + align-items: center; + gap: 0.5rem; + font-size: 0.9rem; + color: var(--text-secondary); + cursor: pointer; +} + +.history-grid { + display: grid; + grid-template-columns: repeat(auto-fill, minmax(280px, 1fr)); + gap: 1rem; + max-height: 500px; + overflow-y: auto; + padding: 0.5rem; +} + +.history-item { + background: rgba(255, 255, 255, 0.03); + border: 1px solid var(--border-color); + border-radius: var(--radius-md); + padding: 1rem; + cursor: pointer; + transition: var(--transition); +} + +.history-item:hover { + background: rgba(255, 255, 255, 0.08); + border-color: var(--border-hover); +} + +.history-item h4 { + font-size: 0.95rem; + margin-bottom: 0.5rem; + overflow: hidden; + text-overflow: ellipsis; + white-space: nowrap; +} + +.history-meta { + font-size: 0.75rem; + color: var(--text-secondary); + display: flex; + justify-content: space-between; + align-items: center; +} + +.history-delete-btn { + background: none; + border: none; + color: var(--text-secondary); + cursor: pointer; + font-size: 0.75rem; + padding: 2px 5px; + border-radius: 4px; + opacity: 0; + transition: var(--transition); + line-height: 1; +} + +.history-item:hover .history-delete-btn { + opacity: 1; +} + +.history-delete-btn:hover { + color: #f87171; + background: rgba(248, 113, 113, 0.15); +} + +.status-badge { + padding: 2px 6px; + border-radius: 4px; + font-size: 0.65rem; + font-weight: 700; +} + +.status-GENERATED { background: rgba(59, 130, 246, 0.2); color: #60a5fa; } +.status-LIVE { background: rgba(16, 185, 129, 0.2); color: #34d399; } + +.modal-input, .modal-textarea { + background: rgba(255, 255, 255, 0.05); + border: 1px solid var(--border-color); + border-radius: var(--radius-md); + color: white; + padding: 0.75rem; + font-family: var(--font-family); + width: 100%; +} + +.modal-textarea { + min-height: 400px; + resize: vertical; + font-family: 'Courier New', Courier, monospace; + font-size: 0.9rem; + line-height: 1.5; +} + +.modal-footer { + margin-top: 2rem; + display: flex; + gap: 1rem; + justify-content: flex-end; +} + /* ── Responsive ────────────────────────────────────────────── */ @media (max-width: 768px) { .navbar { padding: 1rem 1.5rem; }