# Weststar AI APIs

> On-prem AI platform APIs (LLM, vision, speech, face, diagrams, crawling, security, video analytics) at https://api.weststar-ai.com. One bearer token per customer domain (`Authorization: Bearer <token>`) works on every service except marine and live, which have their own tokens. Send it on every call now: most services only log it today, but it will be enforced.

Start with the README. It covers the base URL, auth, quotas, error conventions and which services are publicly routed. Then read the guide for each service you integrate. Every guide follows TEMPLATE.md and uses real captured responses.

## Start here

- [README - integration rules for AI agents](https://api.weststar-ai.com/docs/agents/README.md): base URL, one-token auth and /auth/whoami, quotas, observed error formats, per-service status table
- [TEMPLATE - structure every service guide follows](https://api.weststar-ai.com/docs/agents/TEMPLATE.md): section order and conventions of the per-service guides

## Platform

- [auth](https://api.weststar-ai.com/docs/agents/auth.md): Tokens: check a token with /auth/whoami, health; one bearer token per domain works on every service

## Language

- [llm](https://api.weststar-ai.com/docs/agents/llm.md): OpenAI-compatible chat completions, completions, embeddings and model list over local models
- [chatbot](https://api.weststar-ai.com/docs/agents/chatbot.md): Managed chatbot conversations with memory (not publicly routed yet)
- [agents](https://api.weststar-ai.com/docs/agents/agents.md): Configurable RAG chat agents with document knowledge and tools; token required

## Vision, media and speech

- [visual](https://api.weststar-ai.com/docs/agents/visual.md): Image generation, reference and instruction edits, enhancement, cartoon characters
- [vision](https://api.weststar-ai.com/docs/agents/vision.md): Describe an image: short and detailed explanation plus extracted metadata
- [mockup](https://api.weststar-ai.com/docs/agents/mockup.md): Design plus garment template to realistic product mockup
- [face](https://api.weststar-ai.com/docs/agents/face.md): Face register, recognize, compare, emotion and profile management
- [sound](https://api.weststar-ai.com/docs/agents/sound.md): Whisper speech-to-text, Piper text-to-speech, audio translation (/v1/audio not publicly routed yet)
- [tts](https://api.weststar-ai.com/docs/agents/tts.md): Chatterbox neural text-to-speech, voice library and one-shot voice cloning

## Diagrams

- [architect](https://api.weststar-ai.com/docs/agents/architect.md): Requirement to dark Mermaid architecture diagram HTML; token required
- [orgchart](https://api.weststar-ai.com/docs/agents/orgchart.md): Prompt to organisation chart; token required
- [timeline](https://api.weststar-ai.com/docs/agents/timeline.md): Prompt to project timeline; token required

## Web and data

- [crawler](https://api.weststar-ai.com/docs/agents/crawler.md): Scrape and crawl web pages to clean content (not publicly routed yet)
- [news](https://api.weststar-ai.com/docs/agents/news.md): Topic to array of full news stories with images
- [sitecheck](https://api.weststar-ai.com/docs/agents/sitecheck.md): Website health, SSL and defacement/hijack check (not publicly routed yet)
- [logicdb](https://api.weststar-ai.com/docs/agents/logicdb.md): NoSQL store with logic engine, time travel and natural-language queries (not publicly routed yet)
- [marine](https://api.weststar-ai.com/docs/agents/marine.md): Tides, marine weather and AI boating advisory; uses its own service token

## Security

- [pentest](https://api.weststar-ai.com/docs/agents/pentest.md): Authorized non-destructive security assessment, AI PDF report emailed
- [sentinel](https://api.weststar-ai.com/docs/agents/sentinel.md): LLM-powered web vulnerability assessment pipeline (not publicly routed yet)

## Video and IoT

- [sensetime](https://api.weststar-ai.com/docs/agents/sensetime.md): SenseStudio HTTP-push event ingest, per-event LLM review, analytics
- [inference](https://api.weststar-ai.com/docs/agents/inference.md): CP Monitoring bridge: detection events and HMAC push (not publicly routed yet)
- [live](https://api.weststar-ai.com/docs/agents/live.md): Live CCTV via go2rtc WebRTC/MSE on live.weststar-ai.com; own ?token= gate
- [aiva](https://api.weststar-ai.com/docs/agents/aiva.md): YOLO object detection on live streams with events and annotated snapshots (not publicly routed yet)

## Optional

- [Human-readable docs](https://api.weststar-ai.com/docs/): browsable reference with an interactive Try-it console for signed-in users
- [Agent Guide page](https://api.weststar-ai.com/docs/#/agentguide): how to hand this documentation to an AI agent