From hive
Research and update the LLM API Landscape document with current pricing, providers, free tiers, and recommendations.
npx claudepluginhub skywalking-dev/hiveThis skill uses the workspace's default tool permissions.
Research the current LLM API ecosystem and update `docs/LLM_API_LANDSCAPE.md` with accurate, comprehensive data.
Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.
Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Research the current LLM API ecosystem and update docs/LLM_API_LANDSCAPE.md with accurate, comprehensive data.
Launch three Oraculo research agents in parallel:
Agent A — Frontier & paid providers:
Research ALL current LLM API providers with public pricing as of today.
For each provider: model names, input/output $/MTok, cached pricing, context window, free tier, OAI-compatibility.
Providers to cover: Anthropic, OpenAI, Google, xAI/Grok, DeepSeek, Mistral, Alibaba/Qwen, Baidu/ERNIE, Zhipu AI, Moonshot/Kimi, Cohere, Perplexity, AI21/Jamba, Reka, Meta/Llama hosted options.
Also check: Fireworks AI, SambaNova, Hyperbolic, DeepInfra, Novita AI, Together AI, Amazon Bedrock, Azure OpenAI.
Flag any new providers not in the current doc.
Flag any pricing changes in the last 30 days.
Flag any deprecated or removed models.
Agent B — Free tiers & inference hosts:
Research ALL free-tier and ultra-low-cost LLM inference providers available today.
For each: exact rate limits (RPM, RPD, TPD, TPM), which models are free, permanent vs trial, OAI-compat, tool use, vision, speed.
Cover: Groq, Cerebras, SambaNova, Google Gemini, OpenRouter, SiliconFlow, Cloudflare Workers AI, Together AI, Mistral, Cohere, NVIDIA NIM, HuggingFace, Fireworks, Hyperbolic.
Assess production reliability of each free tier.
Note any recently removed or reduced free tiers.
Agent C — Image & video generation APIs:
Research ALL image and video generation API providers with public pricing.
IMAGE: OpenAI gpt-image-1/DALL-E, Google Imagen 4/Gemini native, BFL FLUX (all variants), Stability AI, Ideogram, Recraft, Leonardo, Replicate, fal.ai, Fireworks, Hyperbolic, Cloudflare Workers AI, Bedrock Titan, Azure DALL-E.
VIDEO: Google Veo 3.x, OpenAI Sora 2, Runway Gen-4, Kling, Pika, Luma, Minimax/Hailuo, ByteDance/Seedance, Wan2.x, CogVideoX, Hunyuan, Bedrock Nova Reel.
For images: price/image, resolution, free tier, key strength, text rendering quality, SVG support.
For video: price/second, max duration, resolution, API availability, free tier.
Note dead ends: Midjourney (no API), Meta Movie Gen (no API), Stability Video (deprecated).
Read docs/LLM_API_LANDSCAPE.md and update ALL sections:
Check which providers in hive/scripts/ already have handlers integrated:
ls /Users/gpublica/workspace/skywalking/hive/scripts/*_handler.*
Update Section 6 integration priority to reflect current handler status (Done vs Pending).
Change the "Last updated" date in the document header to today's date.
**bold** for best-in-class values~~strikethrough~~ for deprecated/legacy modelsThis skill can be run manually (/llm-landscape) or scheduled via cron.
Recommended frequency: monthly (first week of each month).
Pricing changes are infrequent enough that monthly is sufficient; flag breaking changes ad-hoc.