Built this with GPT-6.1 Sol xhigh using just a short prompt. 🤯
From a simple idea to generating an actual 3D project. The coding speed is insane!
Only problem? I hit my usage limit before I could finish. 😂
#GPT6 #Codex #AIcoding #VibeCoding
From a simple idea to generating an actual 3D project. The coding speed is insane!
Only problem? I hit my usage limit before I could finish. 😂
#GPT6 #Codex #AIcoding #VibeCoding
1
0
0
11
Sarah's company policy forbade cloud AI tools. Every demo failed.
Until she found Jarvis.
100% local. Full-stack generation. Zero uploads.
Her proprietary code never left her machine.
#AICoding #DeveloperTools #PrivacyFirst
Until she found Jarvis.
100% local. Full-stack generation. Zero uploads.
Her proprietary code never left her machine.
#AICoding #DeveloperTools #PrivacyFirst
0
0
0
3
The Software Restaurant with Claude: a one-page mental model for AI-assisted software engineering.
You own the restaurant. Claude is the chef. Nothing leaves the kitchen without your taste test.
#ClaudeCode #AIcoding #SoftwareEngineering
You own the restaurant. Claude is the chef. Nothing leaves the kitchen without your taste test.
#ClaudeCode #AIcoding #SoftwareEngineering
0
0
2
6
The AI coding metric I care about most:
How much work remains after the agent says “done”?
Reviewing changes, fixing regressions and checking requirements all count.
I want to measure the time it takes to reach something I can ship.
#AICoding #SoftwareDevelopment
How much work remains after the agent says “done”?
Reviewing changes, fixing regressions and checking requirements all count.
I want to measure the time it takes to reach something I can ship.
#AICoding #SoftwareDevelopment
1
0
0
17
Developers felt 20% faster with AI.
The stopwatch said 19% slower.
4 AI coding myths vs the actual data, and 3 rules that really help ⏱️
Turn "almost right" into done? Offer it as a Task on Anylancer.com
#AIcoding
The stopwatch said 19% slower.
4 AI coding myths vs the actual data, and 3 rules that really help ⏱️
Turn "almost right" into done? Offer it as a Task on Anylancer.com
#AIcoding
0
0
0
1
the 3 MCP concepts that made agent tools click:
• Tools: actions the model executes
• Resources: read-only context
• Prompts: reusable slash workflows
#1 rule: never console.log to stdout in stdio (breaks JSON-RPC).
what custom MCP server did you build first? #AIcoding
• Tools: actions the model executes
• Resources: read-only context
• Prompts: reusable slash workflows
#1 rule: never console.log to stdout in stdio (breaks JSON-RPC).
what custom MCP server did you build first? #AIcoding
0
0
1
35
stop manually copying files into LLMs.
repomix packs your repo into clean XML with file tree, line numbers and token counts.
tree-sitter --compress cuts ~70% tokens on big repos.
github.com/yamadashy/repo…
what do you use to feed context to agents? #AIcoding
repomix packs your repo into clean XML with file tree, line numbers and token counts.
tree-sitter --compress cuts ~70% tokens on big repos.
github.com/yamadashy/repo…
what do you use to feed context to agents? #AIcoding
0
0
1
54
Fresh coding agent leaderboard, 3 things:
1. Claude Code + Opus 5.5 ranks #1 on Artificial Analysis' index (61.7)
2. ~$11.42 per task vs $0.39 for the cheapest frontier pick, 29x
3. It ranks model + harness pairs, not bare models
#AIcoding
Source: artificialanalysis.ai/agents/coding
1. Claude Code + Opus 5.5 ranks #1 on Artificial Analysis' index (61.7)
2. ~$11.42 per task vs $0.39 for the cheapest frontier pick, 29x
3. It ranks model + harness pairs, not bare models
#AIcoding
Source: artificialanalysis.ai/agents/coding
1
0
1
108
Kimi K3.1 Pro: an open-weight coding model narrowly behind Opus 5.5 on Terminal-Bench
kimi k3.1 pro is out and it's open weight.. 65.7% on terminal-bench 4.0 vs opus 5.5 at 66.4% (kimi's own numbers).
Would you switch to Kimi?
#AICoding #OpenSource
kimi k3.1 pro is out and it's open weight.. 65.7% on terminal-bench 4.0 vs opus 5.5 at 66.4% (kimi's own numbers).
Would you switch to Kimi?
#AICoding #OpenSource
0
0
0
23
GitHub's new Teams app turns chat into coding work. I'd separate accepted decisions from suggestions before handing over a thread. Which message should be allowed to change an already-running task?
marcusykim.medium.com/github-copilot… marcusykim.com
#AICoding
marcusykim.medium.com/github-copilot… marcusykim.com
#AICoding
0
0
0
8
asked for a one-line typo fix. got a dependency upgrade, a Dockerfile and a new auth layer.
now every agent task I give ends with "only touch the files you need for this, list anything else as a suggestion" 😅
#AIcoding
now every agent task I give ends with "only touch the files you need for this, list anything else as a suggestion" 😅
#AIcoding
0
0
0
32




















