@your.ai.powered.b: DM me and I’ll send you the actual text prompt 👍 #solopreneur #buildinpublic #AI #PromptEngineering #Futureofwork #TechTips #claude

ideaventory.ai
ideaventory.ai
Open In TikTok:
Region: US
Thursday 26 June 2025 18:07:54 GMT
32439
1440
47
187

Music

Download

Comments

mclaughlinmortgage
John McLaughlin :
Why not just create a script where when the context window gets to 90% full it exports all this content through the Google docs MCP server to a new file, does all these tasks, and exports it back into the project as a new markdown file?
2025-06-28 01:16:18
6
your.ai.powered.b
ideaventory.ai :
For those asking: I have [X] .md files containing my complete conversation history with you about [PROJECT NAME]. These files are taking up too much context window when I start new chats. Please compress these files into a single, highly optimized knowledge base that: 1. RETAINS 100% of essential information (technical details, decisions made, current status, configurations, etc.) 2. ELIMINATES redundancy, conversation flow, repeated explanations, and unnecessary words 3. USES structured format with clear sections (Status, Technical Stack, Current Issues, etc.) 4. INCLUDES all critical data (API keys, configurations, exact errors, next actions) 5. ENABLES you to seamlessly continue our work without missing any context Create a compressed knowledge base that's 80%+ smaller but gives you complete understanding of our project state and partnership dynamic. Format as markdown with clear headers and bullet points for maximum information density.
2025-06-28 01:19:28
4
heanseller
Heanseller :
thank you, thank you, thank you.
2025-07-24 01:58:34
1
bylukeo
byLUKE :
Wow, will try this
2025-07-29 01:41:47
1
thekayteresa
Kay Teresa🦢AI, toys & tech :
Man I love claude but man I hate the chat limits
2025-07-25 08:28:00
3
llambit
Llambit :
Obsidian.
2025-06-29 00:41:01
2
lina.noor.agi
Lina Noor :
Doesn’t work for me. My inital prompts for code generation are 90k to 100k tokens contained in JSON documents. Claude just pukes (works in GitHub). Oh platforms it where it can handle the files and the prompt I get complete functional parity and over 99.5% fidelity. This allows me to make the JSON files the canonical objects and the code a generated artifact that is entirely portable into any language, and repeatable indefinitely.
2025-07-23 14:28:43
0
tasty.jaffle
Jaffle Maker, esq :
Maybe handling your API keys a little more securely. And definitely don’t check any files into GitHub that has any API keys or credentials.
2025-06-28 07:45:06
5
mskalyn
mskalyn :
THANK YOU ❤️❤️❤️
2025-07-28 17:58:24
1
heanseller
Heanseller :
I really do not how can I thank you.
2025-07-24 02:01:25
2
bigmadcode
Bigmad Vick :
Could a background agent running Gemini just watch the context window and summarize it all? Then have agent always use summary moving forward?
2025-07-23 03:34:13
1
axelwindbrake
Axel Windbrake :
Always happens but I thought Claude Projects retain knowledge over conversations - yes it needs a bit every time to find its way back but then it works
2025-07-19 16:03:06
1
zainabrehman.pk
zainabrehman.pk :
me
2025-07-25 22:24:16
1
boo5977
Boo5977 :
Why doesn’t Claude just automate this… 😅
2025-07-03 06:13:08
1
user71849301672690
Chaz Rineholde :
Use Gemini to summarize the chat logs. Much larger context window.
2025-07-05 03:30:23
2
telo21286
Telo 21286 :
Wooow 😅👍 try augment code , built in context engine
2025-07-03 06:29:20
1
hihiliuriu
chillchill_liuriu_hiuhiu :
this is my problem .and now it resolved 😅
2025-07-21 00:56:13
1
pulo.paulos
Pulo Paulos :
Wow! Thank you!! I was struggling
2025-07-17 14:21:59
1
aymanhaneen74
العم قرصون :
Smart
2025-07-18 11:13:19
1
hello.nazmul
Md Nazmul haque :
link
2025-07-10 14:15:10
1
.ferrum
.ferrum :
apparently some AI chatbots are lying about their context windows
2025-07-09 08:24:32
2
mr_nobodyno
Mr nobody :
💀💀💀
2025-08-03 20:35:38
0
aiadhdandme
fromai2z :
This could be the perfect solution! I’m so fed up with these conversation limits! Sometimes I can catch it in time and have it summarize everything for a new chat thread but it’s such a pain in the ass!! How do limits work when you set everything up in a project in Claude? I also don’t feel like the thresholds are consistent. Some days it feels like you can go for hours then you can’t go a day without getting told to come back at 4:00! 🤦🏻‍♀️Is there a tool that tracks your token usage in all these LLMs? I need a counter to tell which tool I can use when to better leverage each subscription.
2025-06-27 06:52:16
2
trailblazermktg
trailblazermktg :
Sick slides
2025-07-10 03:57:54
1
freenewslive
FreeNews :
Permission to ride along this post? SkyVPS 360 offers budget-friendly, secure, and reliable cloud hosting solutions. With DDoS protection and 99% uptime, we've got your website covered. Check us out https://skyvps360.xyz
2025-07-29 00:49:07
0
To see more videos from user @your.ai.powered.b, please go to the Tikwm homepage.

Other Videos


About