Help job seekers know which companies allow AI tools in coding interviews
Problem
More companies are banning or restricting AI assistants in technical interviews, but candidates often don't know the specific rules beforehand, leading to misdirected preparation or on-the-spot violations. Information is scattered and opaque.
Solution
Build a crowdsourced database collecting and verifying company policies on AI tool usage in interviews. Provide search, filtering, and company detail pages; users can submit their own experiences, which are moderated and updated.
Local, fast, and accurate AI image/video authenticity detection for creators and everyday users
Problem
AI-generated images and videos are flooding the internet, but current detection tools either fail on the latest models (like Flux) or have unacceptable false positives on real images. Users need a reliable detector that keeps up with new generative models and runs locally.
Solution
Package state-of-the-art open-source detection models into a simple desktop app. Users drag in images or videos and get a detection result with confidence scores in seconds, with batch processing and all computation done locally for privacy.
Upload a photo and let AI act out an argument for you or your pet
Problem
Users see the 'argument-style performance' meme (e.g., Bad Times AI, Burning Bridges Drake AI) on social media and want to make similar videos with their own photos, but it requires prompt writing and editing, which is a high barrier.
Solution
A simple website where users upload photos (people, cats, dogs) and choose a template (argument, dance, transformation, etc.); AI automatically generates a short video with music and camera work; no prompt or editing needed.
Interactive call graphs for large repos without LLM
Problem
Developers struggle to navigate large codebases and quickly understand function call relationships; existing tools rely on LLMs or cloud, which are slow and raise privacy concerns.
Solution
A local tool that statically analyzes codebases, generates call graphs, and visualizes them; supports fast jump and search, fully offline, no LLM needed.
Compress prompts and history to stretch Claude/ChatGPT subscriptions by 60%
Problem
Users of AI coding assistants hit token limits and must constantly top up or switch accounts. Someone built Mixdog to cut token usage by 63%, but similar solutions are not widespread.
Solution
A middleware tool that automatically compresses conversation history and prompts, removing redundancy while preserving key info, reducing token usage; supports major AI clients and APIs.
One-click security scan for prompt injection and supply-chain risks in AI-generated apps
Problem
Users hesitate to install vibe-coded apps, worrying about prompt injection or hidden viruses. These apps often lack human code review, creating high security risks.
Solution
A scanning tool where users upload or link to app code/binaries; it analyzes dependencies, prompt injection patterns, suspicious network calls, and gives a security score with fix suggestions; can integrate into CI or browser extension.
Self-creating email inboxes that lock to one sender and bounce all spam
Problem
Users are overwhelmed by spam and marketing emails, and regular inboxes are hard to manage. Someone built an email app that only accepts one sender per inbox, but no product serves mainstream users yet.
Solution
An email service that creates a new inbox on-the-fly when you use any prefix@yourdomain; each inbox locks to the first sender domain (or email address) and bounces everyone else; temporary inboxes expire after a set time (e.g., 24h, 7 days) and bounce all future mail.
Multi-engine inline translation and OCR for reading and writing workflows
Problem
A developer open-sourced a cross-platform inline translation tool, showing that selecting text in a browser or editor, pressing a hotkey, and seeing the translation in a floating window near the cursor is still a high-frequency need. Users also want side-by-side comparison across translation providers, screenshot OCR, and keys that are not stored in plaintext config files.
Solution
A cross-platform inline translation tool with multi-provider side-by-side comparison, screenshot region OCR, silent replace back into the selection, history and favorites, and keys stored in the system keychain.
A lightweight tool for recording, comparing, and reviewing physical store locations
Problem
A developer who wanted to open a physical store found it inconvenient to check locations on Amap, jot down ideas, and store on-site photos, so he built his own site-selection tool with 1/3/5 km POI data, population heatmaps, and AI evaluation. This shows ordinary founders lack a handy tool for site research; information is scattered across map apps and photo albums.
Solution
Combine map selection, location notes, on-site photos, and a simple evaluation template into a lightweight web app where users can quickly compare candidate spots, log thoughts and photos from each visit, and generate a comparison report with map screenshots.
Orchestrate Claude Code, Codex, and other CLI agents in one board with clear progress
Problem
A developer asked what people do while waiting for their copilot to finish—programming has become managerial, flow is gone, yet you still juggle multiple agents and features. Others have built open-source desktop tools to put CLI agents into one crew, but conflicts and visibility when multiple agents work together remain painful.
Solution
A local desktop board that aggregates terminal output from Claude Code, Codex, Cursor, and other CLI agents, showing each agent's current task, blockers, and last activity, with support for adding/removing roles mid-task and relaying messages.
Generate appeal letters and timelines for indie founders whose funds are frozen by payment platforms
Problem
A developer reported that Wise closed his company accounts and held roughly $22,000 plus 3,400 PLN for 27 days, then went silent, saying only that it could not share the specific reason. Indie founders and small business owners in this situation often do not know whom to appeal to, what evidence to prepare, or how to keep a clean timeline.
Solution
Users import emails, notices, and account statements from the payment platform, and the tool auto-generates a timestamped event timeline, draft appeal letters, and regulatory complaint templates (e.g., financial ombudsman, state regulators), plus deadline reminders for each channel.
Push relevant open government tenders to small businesses before deadlines hit
Problem
There are over 250,000 open government tenders worldwide, scattered across national portals in different formats and languages. SME owners and solo consultants cannot possibly browse them all, so they miss tenders that fit their capabilities perfectly—or find them after the bidding window closes.
Solution
Aggregate open government tender data, run semantic matching and deduplication against a company profile (industry, region, qualifications, keywords), and deliver relevant tenders daily via email or RSS with deadlines, budget ranges, and source links.
Parse customs tax bills and generate review requests when fees seem off
Problem
A user shipping electronics from Taiwan to mainland China was charged 642 RMB in tax by SF Express without a clear breakdown, and did not know who to ask for the official bill or how to request a review. Cross-border parcel taxation is opaque and appeals are cumbersome.
Solution
A small tool where you upload a waybill or tax bill screenshot; it OCRs item value, tax rate, and due amount, compares with actual charge, generates a calculation explanation and a review request template, plus steps to query customs or courier.
Browser-based one-click fake data, API fixtures, and seed scripts
Problem
Developers often need to fabricate user profiles, credit card numbers, and database seed data for testing. Doing it manually is slow and unrealistic, while online tools raise data leakage concerns or are clunky.
Solution
A purely browser-based mock data generator with custom fields (name, address, card number, UUID, etc.), export to JSON/CSV/SQL seed scripts, and savable presets. All data is generated locally, never uploaded.
Ask once and send to multiple AIs, compare answers side by side, no API keys
Problem
When using AI for research, you must copy the same question into multiple tabs and switch back and forth to compare. Users want to compare using already logged-in accounts instead of copy-pasting or configuring APIs.
Solution
A browser extension that opens multiple AI chat pages in split view, sends one input to all, and shows answers side by side. It reuses existing logged-in sessions, requires no API keys, and supports prompt templates with variables.
Manage multi-agent, multi-project work around tasks, not agents, with handoff
Problem
Multiple coding agents each keep their own work logs, so when one agent takes over from another, context must be re-explained. When juggling several projects, it is hard to know which tasks are done, in progress, or need human intervention without manually syncing across sessions.
Solution
A local-first task board where each task binds its context (goal, constraints, progress, artifacts). When a session ends, export a context package; a new agent imports it and continues. A board aggregates agent status and supports manual task state and intervention flags.
Always-visible quota and approval requests for AI coding agents on the Mac notch or edge
Problem
When running Claude Code, Codex, and other coding agents in parallel, you have to switch windows constantly to see which quota is running out and which agent is waiting for your approval. Developers report this breaks flow and that quota warnings come too late, causing rate-limit surprises at critical moments.
Solution
A tiny always-on Mac menu bar/notch app that aggregates quota usage, reset times, burn rate, and balance across agents, and surfaces pending approval requests with the command and file diff so you can approve or deny in one click. It alerts before quota runs out based on burn-rate projection.
Interactive audio ink-wash map explaining geography and history in ancient Chinese poetry
Problem
Parents teaching children ancient poems find that kids have no concept of place names or poets' journeys, as traditional text or static maps fail to engage. One developer spent a month making an audio ink-wash map, but ordinary parents cannot replicate that effort.
Solution
An online tool that selects common ancient poems and automatically generates an interactive map with audio narration. Users pick a poem, and the map dynamically shows the poet's footsteps and related locations, with ink-wash style visuals and a timeline. Audio explains background stories and geographic changes.
One-click export and migration of AI chat histories to prevent platform lock-in
Problem
A user lost nearly a year of chat history and research data when Zhipu's overseas site suddenly blocked Chinese users. Similarly, other platforms may change terms or impose regional restrictions anytime, and users often lack local backups, causing critical data to vanish. Users feel as distressed as losing a phone.
Solution
A desktop tool that exports conversation history from multiple AI chat platforms (ChatGPT, Claude, Gemini, Zhipu, etc.) into a universal format (JSON/Markdown) and supports importing into other platforms or local knowledge bases. It also offers scheduled automatic backups to ensure a local copy always exists.
LLM API capacity status dashboard with alerts and fallback suggestions
Problem
Developers using services like Codex often encounter 'Selected model is at capacity' errors and must manually try other models. Users complain this disrupts workflow, especially for paid accounts, and there's no easy way to check model availability or auto-switch.
Solution
A status monitoring service that probes major LLM APIs (OpenAI, Anthropic, Google, DeepSeek, etc.) in real-time for capacity, via simulated requests or public status pages, and presents an availability dashboard. Developers can subscribe to specific models, get notified when capacity is low, and receive fallback suggestions.