My take after 3 months with AugmentCode:
- Best agent I’ve tried so far—quite pricey (haven’t tried Claude Code since likely it more expensive).
- Great for planning, brainstorming, and scaffolding. Context gathering is solid, even with large codebases.
- Still kinda risky for full implementation—especially at scale.
My take after 3 months with AugmentCode:
- Best agent I’ve tried so far—quite pricey (haven’t tried Claude Code since likely it more expensive).
- Great for planning, brainstorming, and scaffolding. Context gathering is solid, even with large codebases.
- Still kinda risky for full implementation—especially at scale.
@txtdarisugab Gua mungkin bias, tapi AugmentCode also pretty ok, likely bisa di combo sama tab autocompletenya cursor free version atau pake with pure vscode juga bisa since augment udah ada autocomplete juga (ga se agresif cursor tapi)
https://t.co/f1EgwHNZIN
My take after 3 months with AugmentCode:
- Best agent I’ve tried so far—quite pricey (haven’t tried Claude Code since likely it more expensive).
- Great for planning, brainstorming, and scaffolding. Context gathering is solid, even with large codebases.
- Still kinda risky for full implementation—especially at scale.
Bottom line: Quality input → quality output.
I might not be using the agent optimally yet—curious how others are approaching it.
How would you do it differently?
- I usually need to refine both the plan and implementation to deal with overengineering. Small iterations help a lot.
- A lot of time I use it for planning/brainstorming—love how it gathers context (especially combned with sequential thinking mcp + proper user guidelines).
- When I want to avoid risk during implementation, I stick to 1 file/thread. Might costs more, but gives me control.
Lewat trained knowledge dari modelnya dan context dari user. MCP basically gunanya untuk enable users and community (and some 3rd party company) to defined the integration with their own data source to expand the model knowledge. MCP meant to solve the problem where the model gabisa akses very specific knowledge yang user mau.
This is my contribution to the dev community. Hopfully bisa bantu buat land you to a better career - 100% open source & free!
Try it out: npm: https://t.co/cISnDKZuJp github: https://t.co/nt9yH03JHW
That's why I built gitlog-author with:
- Auto-generate contribution reports yang lebih enak dilihat
- Productivity insights & patterns
- Detailed impact analysis
- Directory filtering untuk scoping di directory tertentu
- Date range filtering untuk fokus di tanggal tertentu
So far toolsnya coock untuk bantu:
- Developers yang mau performance review
- Tech leads yang butuh clear view of team progress
- Teams yang perlu proper handover docs
- Yang mau #KaburAjaDulu