mirror of
https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git
synced 2026-06-18 15:29:36 +00:00
4bb2442313
Complete AI-powered personal brand automation for Sami Assiri.\n\n7 agents: LinkedIn, Email, Social Media, WhatsApp, CV Optimizer, Content Strategist, Opportunity Scout.\nInfra: FastAPI + APScheduler + Docker + Ollama/Groq LLM + GitHub Pages landing page.\n83 files, ~10K lines. Cost: $0-5/month.
234 lines
7.8 KiB
Python
234 lines
7.8 KiB
Python
"""Calendar planner -- generates a 7-day content plan aligned with brand strategy."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import logging
|
|
from datetime import date, datetime, timedelta, timezone
|
|
from typing import Any
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
# Days of the week when content is posted (0=Mon ... 6=Sun)
|
|
# From schedule.yaml: Sun/Tue/Thu -- ISO weekday: Sun=7, Tue=2, Thu=4
|
|
# Python date.isoweekday(): Mon=1, Tue=2, Wed=3, Thu=4, Fri=5, Sat=6, Sun=7
|
|
_POSTING_DAYS_ISO = {7, 2, 4} # Sunday, Tuesday, Thursday
|
|
|
|
_SYSTEM_PROMPT = """\
|
|
You are a LinkedIn content strategist for a Field Services Engineer specializing
|
|
in Smiths Detection airport security equipment, based in Riyadh, Saudi Arabia.
|
|
|
|
Create a 7-day content calendar. Posts are scheduled for Sunday, Tuesday, and Thursday.
|
|
The other days are for engagement-only (likes, comments, networking).
|
|
|
|
For each posting day, provide:
|
|
- "date": ISO date string (YYYY-MM-DD)
|
|
- "pillar": one of the content pillars from the strategy
|
|
- "topic": specific topic / angle for the post
|
|
- "platform": "linkedin" (primary) or "twitter"
|
|
- "suggested_hook": the opening line / hook for the post (1-2 sentences)
|
|
- "hashtags": list of 3-5 relevant hashtags
|
|
- "content_type": "text", "carousel", "poll", "video_script", or "article"
|
|
|
|
For non-posting days, include an engagement-only entry:
|
|
- "date": ISO date string
|
|
- "pillar": "engagement"
|
|
- "topic": "Network engagement & community interaction"
|
|
- "platform": "linkedin"
|
|
- "suggested_hook": ""
|
|
- "hashtags": []
|
|
- "content_type": "engagement"
|
|
|
|
Ensure variety across pillars and content types throughout the week.
|
|
Return ONLY a valid JSON array of 7 objects (one per day).
|
|
"""
|
|
|
|
|
|
async def generate_weekly_calendar(
|
|
llm_client: Any,
|
|
brand_profile: dict,
|
|
content_strategy: dict,
|
|
trends: list[dict] | None = None,
|
|
) -> list[dict]:
|
|
"""Generate a 7-day content plan starting from the next Sunday.
|
|
|
|
Parameters
|
|
----------
|
|
llm_client:
|
|
An :class:`LLMClient` instance.
|
|
brand_profile:
|
|
Parsed ``brand_profile.yaml``.
|
|
content_strategy:
|
|
Parsed ``content_strategy.yaml``.
|
|
trends:
|
|
Optional list of trending topics from :func:`analyze_trends`.
|
|
|
|
Returns
|
|
-------
|
|
list[dict]
|
|
Seven entries, one per day, each with ``date``, ``pillar``, ``topic``,
|
|
``platform``, ``suggested_hook``, and metadata.
|
|
"""
|
|
# Calculate the start of next week (next Sunday)
|
|
today = date.today()
|
|
days_until_sunday = (7 - today.isoweekday()) % 7
|
|
if days_until_sunday == 0:
|
|
days_until_sunday = 7 # If today is Sunday, plan for next week
|
|
week_start = today + timedelta(days=days_until_sunday)
|
|
|
|
week_dates = [week_start + timedelta(days=i) for i in range(7)]
|
|
|
|
# Build the pillar descriptions for the prompt
|
|
pillars = content_strategy.get("content_pillars", [])
|
|
pillar_text = "\n".join(
|
|
f"- {p['id']}: {p.get('name_en', '')} -- {p.get('description', '')}"
|
|
for p in pillars
|
|
)
|
|
|
|
# Format trends
|
|
trends_text = ""
|
|
if trends:
|
|
trends_text = "Trending topics to consider:\n" + "\n".join(
|
|
f"- {t.get('topic', '')} (relevance: {t.get('relevance', 'medium')}, pillar: {t.get('pillar', '')})"
|
|
for t in trends[:8]
|
|
)
|
|
|
|
personal = brand_profile.get("personal", {})
|
|
user_prompt = f"""\
|
|
Generate a 7-day content calendar for the week of {week_start.isoformat()} to {week_dates[-1].isoformat()}.
|
|
|
|
Professional context:
|
|
- Name: {personal.get('name_en', '')}
|
|
- Role: {personal.get('title_en', '')}
|
|
- Company: {brand_profile.get('employment', {}).get('current', {}).get('company', '')}
|
|
- Location: {personal.get('location_en', '')}
|
|
|
|
Content pillars:
|
|
{pillar_text}
|
|
|
|
Posting schedule: Sunday, Tuesday, Thursday
|
|
Engagement-only days: Monday, Wednesday, Saturday, Friday
|
|
|
|
Tone: {content_strategy.get('tone', {}).get('style', 'professional_approachable')}
|
|
Primary language: Arabic (with English for technical/international content)
|
|
|
|
{trends_text}
|
|
|
|
Week dates:
|
|
{chr(10).join(f'- {d.isoformat()} ({d.strftime("%A")})' for d in week_dates)}
|
|
|
|
Return ONLY a valid JSON array of 7 objects.
|
|
"""
|
|
|
|
response = await llm_client.generate(
|
|
prompt=user_prompt,
|
|
system_prompt=_SYSTEM_PROMPT,
|
|
temperature=0.6,
|
|
max_tokens=2500,
|
|
)
|
|
|
|
calendar = _parse_calendar_response(response.text, week_dates)
|
|
logger.info("Weekly calendar generated: %d entries", len(calendar))
|
|
return calendar
|
|
|
|
|
|
def _parse_calendar_response(
|
|
text: str,
|
|
week_dates: list[date],
|
|
) -> list[dict]:
|
|
"""Parse the LLM response into a calendar list, with fallback generation."""
|
|
cleaned = text.strip()
|
|
|
|
# Strip markdown code fences
|
|
if cleaned.startswith("```"):
|
|
first_newline = cleaned.index("\n")
|
|
cleaned = cleaned[first_newline + 1 :]
|
|
if cleaned.endswith("```"):
|
|
cleaned = cleaned[: -len("```")].rstrip()
|
|
|
|
try:
|
|
parsed = json.loads(cleaned)
|
|
if isinstance(parsed, list):
|
|
entries = parsed
|
|
elif isinstance(parsed, dict) and "calendar" in parsed:
|
|
entries = parsed["calendar"]
|
|
else:
|
|
entries = [parsed]
|
|
except json.JSONDecodeError:
|
|
logger.warning("Failed to parse calendar JSON; generating fallback")
|
|
return _generate_fallback_calendar(week_dates)
|
|
|
|
# Validate and normalize entries
|
|
normalized: list[dict] = []
|
|
for entry in entries:
|
|
normalized.append(
|
|
{
|
|
"date": entry.get("date", ""),
|
|
"pillar": entry.get("pillar", "engagement"),
|
|
"topic": entry.get("topic", ""),
|
|
"platform": entry.get("platform", "linkedin"),
|
|
"suggested_hook": entry.get("suggested_hook", ""),
|
|
"hashtags": entry.get("hashtags", []),
|
|
"content_type": entry.get("content_type", "text"),
|
|
}
|
|
)
|
|
|
|
# Ensure we have exactly 7 entries (pad with engagement days if needed)
|
|
while len(normalized) < 7:
|
|
idx = len(normalized)
|
|
if idx < len(week_dates):
|
|
d = week_dates[idx]
|
|
else:
|
|
d = week_dates[-1] + timedelta(days=idx - len(week_dates) + 1)
|
|
normalized.append(
|
|
{
|
|
"date": d.isoformat(),
|
|
"pillar": "engagement",
|
|
"topic": "Network engagement & community interaction",
|
|
"platform": "linkedin",
|
|
"suggested_hook": "",
|
|
"hashtags": [],
|
|
"content_type": "engagement",
|
|
}
|
|
)
|
|
|
|
return normalized[:7]
|
|
|
|
|
|
def _generate_fallback_calendar(week_dates: list[date]) -> list[dict]:
|
|
"""Generate a basic fallback calendar when LLM parsing fails."""
|
|
_fallback_pillars = [
|
|
"tech_insights",
|
|
"engagement",
|
|
"field_life",
|
|
"engagement",
|
|
"industry_news",
|
|
"engagement",
|
|
"engagement",
|
|
]
|
|
_fallback_topics = [
|
|
"Weekly airport security technology insight",
|
|
"Network engagement & community interaction",
|
|
"A day in the life of a Field Services Engineer",
|
|
"Network engagement & community interaction",
|
|
"Industry news commentary and analysis",
|
|
"Network engagement & community interaction",
|
|
"Network engagement & community interaction",
|
|
]
|
|
|
|
entries: list[dict] = []
|
|
for i, d in enumerate(week_dates[:7]):
|
|
is_posting_day = d.isoweekday() in _POSTING_DAYS_ISO
|
|
entries.append(
|
|
{
|
|
"date": d.isoformat(),
|
|
"pillar": _fallback_pillars[i] if is_posting_day else "engagement",
|
|
"topic": _fallback_topics[i] if is_posting_day else "Network engagement & community interaction",
|
|
"platform": "linkedin",
|
|
"suggested_hook": "",
|
|
"hashtags": [],
|
|
"content_type": "text" if is_posting_day else "engagement",
|
|
}
|
|
)
|
|
return entries
|