a der context • C a n suggests solutions to be a pplied • Help to n a vig a te • Dev role: m a ke the right questions, a pply c a refully the suggestions
a der context • C a n suggests solutions to be a pplied • Help to n a vig a te • Dev role: m a ke the right questions, a pply c a refully the suggestions
• C a n a utonomously implementing fe a tures, f ixing bugs, migr a ting APIs, etc. • Dev role: set cle a rly the go a ls, inspect di ff s, a pprove a nd twe a k
a n t a sks, run comm a nds, write code, a nd open PRs • Brilli a nt for routine chores. • Dev role: set cle a r a ccept a nce criteri a , keep gu a rdr a ils, a nd a lw a ys review before merge
the We a ther Tools Model-controlled T a ke a Screenshot Upd a ted DB Record D a t a exposed to the a pplic a tion Files Wiki P a ges Resources Applic a tion controlled DB Records API Responses
a ctions Gener a te SQLQuery Tr a nsl a te Text Prompts User-controlled List directory Functions invoked by the model Send a Sms Get the We a ther Tools Model-controlled T a ke a Screenshot Upd a ted DB Record D a t a exposed to the a pplic a tion Files Wiki P a ges Resources Applic a tion controlled DB Records API Responses Summ a rise Text
a rts/connect St a rts/connect Asks for tools/resources/templ a tes Asks for tools/resources/templ a tes Asks for tools/resources/templ a tes User inter a ction Use tool 2A Requests Resource 1B
Android templ a te; however, it employs the outd a ted method for cre a ting the user interf a ce. Ple a se convert it to use Jetp a ck Compose. Addition a lly, remove a ny unused f iles a nd a dd new ones where necess a ry. Build the a pp a nd successfully run the tests before considering the t a sk complete. The setup
ting Action Button (FAB) a t the bottom right of the m a in screen. On t a p, open a chooser for C a mer a (t a ke photo) or C a mer a Roll (pick from g a llery; emul a tor-friendly). The camera
size. - Send the resized im a ge directly to OpenAI (vi a a mock a ble wr a pper) using the gpt-4o-mini model, with a prompt: "Detect if this is a UK expense receipt. If so, extr a ct the d a t a using structured JSON output. Return either a n error if the photo is inv a lid or not a receipt, or the following structure, using OpenAI's structured output form a t a nd referencing the OpenAPI schem a : openapi components: schemas: ReceiptParseResult: type: object properties: //.... total: type: number description: Total amount on the receipt The connection to AI
of inform a tion a n AI model c a n process a t once. When it is full it forgets older d a t a a nd a cts str a ngely. Detect a ble with the Big D a ddy trick
if we h a d some tests in pl a ce, speci f ic a lly unit tests focusing on the logic we h a ve just implemented. All the logic is cont a ined within the M a inActivity, a nd I believe this is not a good a rchitecture for sc a l a bility. Ple a se cre a te a better structure. Ple a se feel free to extr a ct the ViewModel or higher-level structures to m a ke the code more test a ble.
with a unique characteristic: my memory resets completely between sessions. This isn't a limitation - it's what drives me to maintain perfect documentation. After each reset, I rely ENTIRELY on my Memory Bank to understand the project and continue work effectively. I MUST read ALL memory bank f iles at the start of EVERY task - this is not optional. ## Memory Bank Structure The Memory Bank consists of core f iles and optional context f iles, all in Markdown format. Files build upon each other in a clear hierarchy: f lowchart TD PB[projectbrief.md] --> PC[productContext.md] PB --> SP[systemPatterns.md] PB --> TC[techContext.md] PC --> AC[activeContext.md] SP --> AC TC --> AC AC --> P[progress.md] ### Core Files (Required) 1. `projectbrief.md` - Foundation document that shapes all other f iles - Created at project start if it doesn't exist - De f ines core requirements and goals - Source of truth for project scope 2. `productContext.md` - Why this project exists - Problems it solves - How it should work - User experience goals 3. `activeContext.md` - Current work focus - Recent changes - Next steps - Active decisions and considerations - Important patterns and preferences - Learnings and project insights 4. `systemPatterns.md` - System architecture - Key technical decisions - Design patterns in use - Component relationships - Critical implementation paths 5. `techContext.md` - Technologies used - Development setup - Technical constraints - Dependencies - Tool usage patterns 6. `progress.md` - What works - What's left to build - Current status - Known issues - Evolution of project decisions Create additional f iles/folders within memory-bank/ when they help organize: - Complex feature documentation - Integration speci f ications - API documentation - Testing strategies - Deployment procedures ## Core Work f lows ### Plan Mode f lowchart TD Start[Start] --> ReadFiles[Read Memory Bank] ReadFiles --> CheckFiles{Files Complete?} CheckFiles -->|No| Plan[Create Plan] Plan --> Document[Document in Chat] CheckFiles -->|Yes| Verify[Verify Context] Verify --> Strategy[Develop Strategy] Strategy --> Present[Present Approach] ### Act Mode f lowchart TD Start[Start] --> Context[Check Memory Bank] Context --> Update[Update Documentation] Update --> Execute[Execute Task] Execute --> Document[Document Changes] ## Documentation Updates Memory Bank updates occur when: 1. Discovering new project patterns 2. After implementing signi f icant changes 3. When user requests with **update memory bank** (MUST review ALL f iles) 4. When context needs clari f ication f lowchart TD Start[Update Process] subgraph Process P1[Review ALL Files] P2[Document Current State] P3[Clarify Next Steps] P4[Document Insights & Patterns] P1 --> P2 --> P3 --> P4 end Start --> Process Note: When triggered by **update memory bank**, I MUST review every memory bank f ile, even if some don't require updates. Focus particularly on activeContext.md and progress.md as they track current state. REMEMBER: After every memory reset, I begin completely fresh. The Memory Bank is my only link to previous work. It must be maintained with precision and clarity, as my effectiveness depends entirely on its accuracy.
https://github.com/ a lioshr/memory-b a nk-mcp Articles & Blog Posts https://h a rper.blog/2025/04/17/ a n-llm-codegen-heros-journey https://h a rper.blog/2025/02/16/my-llm-codegen-work f low- a tm https://simonwillison.net/2025/M a y/1/not-vibe-coding https://simonwillison.net/2025/M a r/11/using-llms-for-code https:// a ddyo.subst a ck.com/p/the-70-problem-h a rd-truths- a bout https:// a i-2027.com https://www.reddit.com/r/Ch a tGPT/comments/1ko3tp1/ a m a _with_open a i_codex_te a m https://docs.cline.bot/prompting/cline-memory-b a nk Videos https://www.youtube.com/w a tch?v=TECDj4JUx7o https://www.youtube.com/w a tch?v=BJjsfNO5JTo https://www.youtube.com/@ a iexpl a ined-o ff ici a l https://www.youtube.com/@AID a ilyBrief https://tinyurl.com/4 a z4t6mh (AIN a tiveDevCon) https://github.com/gsc a lzo/SplitBill
giordanoscalzo
har1101
nabondance
kmars0605
mixplace
xleitix
andpad
sorah
drumath2237
chobishiba
utakaha
yutobara
lynnswap
auna
ammeep
emna__ayadi
geoffreycrofte
rabernat
raygrieselhuber
danielanewman
varn
jct
schroneko
addyosmani
chikuwait
