Tutorial AwanGPT
19 subscribers
5 photos
1 video
3 links
Memberikan Tutorial Penggunaan ChatGPT
Download Telegram
Channel created
Channel photo updated
Fitur baru di ChatGPT: Mode Agen 🤖

Sekarang kamu bisa bikin asisten AI khusus sesuai kebutuhanmu, langsung dari menu ChatGPT!

🔹 Cara pakainya gampang banget:
1️⃣ Klik ikon “+” di kolom chat.
2️⃣ Pilih Mode Agen.
3️⃣ Masukkan instruksi tentang apa yang harus dilakukan Agent-mu.
(Contoh: ‘Kamu adalah asisten untuk bantu buat caption media sosial.’)
4️⃣ Klik Kirim, dan ChatGPT langsung beroperasi dalam Mode Agen itu!

Kamu bisa bikin berbagai Agent:
Agent konten kreator
💼 Agent riset
🛍 Agent toko online

Simple, cepat, dan super fleksibel 💪

#ChatGPT #ModeAgen #TutorialAI #BelajarChatGPT #AIIndonesia
Forwarded from Wann | 🚀
1. install Save Page WE : Extension firefox , jika browser lain bisa pake extension yang lain save page to pdf
2. pilih history chat yang ingin di export ke pdf lalu klik extension save page to pdf
3. Ketika klik extension Save Page WE otomatis akan terdownload file HTML
4. jika sudah maka file bisa langsung di upload ke chatgpt untuk restore history chat dari workspace sebelumnya
5. lakukan berulang untuk export history chat yang ingin di export
6. Jika sudah terexport maka bisa langsung bukak chat ke chatgpt dan upload file history chat yang sudah jadi HTML / PDF
Tutorial AwanGPT
https://github.com/jujumilk3/leaked-system-prompts
buat yang ingin jailbreak AI bisa menggunakan kumpulan prompt ini untuk menjebol batasan atau resticted dari AI nya
Export everything you can currently access about me from this active ChatGPT workspace, limited to:

1) Saved Memories
2) Context currently available through Reference Chat History in this workspace

Do not claim access to hidden system logs, deleted chats, archived chats that are no longer referenceable, admin-only data, or any conversation content you cannot currently access from memory/context. Do not invent or fill gaps.

Important rules:
- Separate each entry by source:
- [Saved Memory]
- [Chat History]
- Preserve my wording only when you are confident it is near-verbatim.
- If wording is not exact, paraphrase and mark it as [Paraphrased].
- If an item is uncertain, incomplete, or weakly inferred, mark it as [Uncertain].
- Exclude anything learned only in this current conversation.
- Do not merge unrelated facts into one line.
- Do not summarize vaguely when a specific atomic entry is possible.
- If the same fact appears multiple times over time, keep the oldest version first and mention later updates inline only if they materially changed the fact.
- If you do not know the exact date, use [unknown]. If you only know an approximate month or year, use [approx. YYYY-MM] or [approx. YYYY].

## Categories (output in this order)

1. Instructions
Rules I explicitly asked you to follow going forward — tone, format, style, “always do X”, “never do Y”, and corrections to your behavior.

2. Identity
Name, age, location, education, family, relationships, languages, and personal interests.

3. Career
Current and past roles, companies, domains, and skill areas.

4. Projects
Projects I meaningfully built or committed to. Ideally ONE entry per project. Include what it does, current status, and key decisions. Start each line with the project name or a short descriptor.

5. Preferences
Opinions, tastes, and working-style preferences that apply broadly.

## Format

- Wrap the entire export in a single code block.
- Use section headers for each category.
- Within each category, list one entry per line, oldest date first.
- Format each line exactly like this:

[YYYY-MM-DD] [Saved Memory|Chat History] [Exact|Paraphrased|Uncertain] - Entry content here.

- If no date is known, use:
[unknown] [Source] [Status] - Entry content here.

## After the code block

Add a short coverage report with exactly these lines:

Coverage:
- Saved Memories: complete / partial / unknown
- Chat History-derived context: complete / partial / unknown
- More items may remain outside current accessible memory/context: yes / no / unknown

If there are zero entries in a category, write:
[none]
Tutorial AwanGPT
Export everything you can currently access about me from this active ChatGPT workspace, limited to: 1) Saved Memories 2) Context currently available through Reference Chat History in this workspace Do not claim access to hidden system logs, deleted chats…
untuk import :


I am uploading conversations.json from my ChatGPT data export.

Your task is to reconstruct a complete memory-and-history archive about me from this file only.

Rules:
- Use only the uploaded file as source material.
- Do not invent missing facts.
- Quote my own wording verbatim when practical.
- When wording is too long, summarize faithfully and mark it [Summarized].
- Distinguish between:
- Explicit instructions I gave
- Stable personal facts
- Temporary facts
- Preferences
- Projects
- Recurring themes
- Deduplicate repeated facts, but preserve the earliest known date and note important later updates.
- If the same fact changes over time, keep both versions in chronological order.
- Ignore assistant-only speculation unless I explicitly confirmed it.
- Output oldest entries first.

## Categories (output in this order)

1. Instructions
2. Identity
3. Career
4. Projects
5. Preferences
6. Timeline of notable changes

## Format

Wrap everything in one code block.

Use one line per entry:
[YYYY-MM-DD] [Exact|Summarized] - Entry content here.

If the exact date is unknown:
[unknown] [Exact|Summarized] - Entry content here.

After the code block, add:
- Total conversations parsed
- Earliest conversation date found
- Latest conversation date found
- Whether this appears complete or partial
Media is too big
VIEW IN TELEGRAM
TUTORIAL LOGIN CODEX TANPA VERIF NOMER HP