skills.rest
waiyanphyo999
waiyanphyo999COMMUNITY
Explore all skills by this author →
GitHub

builtin-llm-models

Documents the built-in LLM proxy catalog, request shapes, and batch calling patterns.

1|Updated Jul 17, 2026
Search Tags:#prompt engineering#api#llm#batch processing#openai#structured output#model catalog
One-click install
npx skills add https://github.com/waiyanphyo999/telegram-bot-deploy --skill builtin-llm-models-waiyanphyo999
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: builtin-llm-models
Source: https://github.com/waiyanphyo999/telegram-bot-deploy/tree/main/skills/builtin-llm-models
Command: npx skills add https://github.com/waiyanphyo999/telegram-bot-deploy --skill builtin-llm-models-waiyanphyo999

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires openai, and includes scripts (resource) components.

What problem does it solve? Calling the built-in LLM proxy incorrectly wastes tokens and causes silent failures: model IDs and prices change, each provider family (GPT, Claude, Gemini) requires different thinking/reasoning parameters, and wrong max-token settings produce null or truncated output. This Skill provides the verified request shapes and a batch driver so bulk LLM jobs work the first time. ## Core Features & Use Cases - Live Model Catalog: Instructions to query the /models endpoint for current IDs, pricing, and capabilities instead of relying on stale snapshots. - Family-Specific Request Shapes: Exact parameters for GPT reasoning effort, Claude thinking budgets, Gemini max_tokens, JSON-schema structured output, vision, and tool calling. - Batch Driver Script: scripts/llm_batch.py runs concurrent, retrying bulk jobs (summarization, translation, classification, extraction) writing JSONL results. - Use Case: You need to summarize 500 documents in the sandbox. Fetch the live catalog, pick gpt-5-mini, and run llm_batch.py with a prompt template to get JSONL output with token usage per item. ## Quick Start Ask the assistant to fetch the live model catalog and run a batch summarization over your input file using the llm_batch.py script with gpt-5-mini.

Frequently Asked Questions about builtin-llm-models

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I call the built-in LLM proxy from a Python script?▼

Use the openai Python SDK with OpenAI(), which auto-reads the preconfigured OPENAI_API_KEY and OPENAI_API_BASE environment variables. Call client.chat.completions.create with a model ID from the live /models catalog.

How do I run batch LLM jobs like bulk summarization or translation?▼

Use the bundled scripts/llm_batch.py driver with an input file of lines or a JSON array, a prompt template containing {input}, and an output JSONL path. It runs concurrent workers with retries and records token usage per item.

Which built-in model should I use for cheap bulk tasks?▼

gpt-5-mini is the default cheap workhorse for most bulk jobs at 0.25/2.00 USD per million tokens. Drop to gpt-5-nano for very large volumes with loose quality tolerance, and skip thinking parameters for simple tasks.

Why does Gemini return content null with finish_reason length?▼

This happens when max_completion_tokens is passed to Gemini models on the proxy. Switch to max_tokens (for example 16384) or omit the parameter entirely, since Gemini reasons by default without explicit thinking parameters.

Why does Claude return a 400 Bad Request with thinking enabled?▼

Claude requires max_tokens to be strictly greater than the thinking budget_tokens value. For example, budget_tokens of 2048 requires max_tokens of at least 2049 in the request.

Does the built-in LLM proxy support streaming responses?▼

No, the proxy reports supports_streaming as false for all models and invokeLLM does not expose a stream parameter. Streaming requires modifying the webdev helper to send stream true and parse the SSE response manually.

Related Skills

Explore More Software Engineering Skills→
gtbaukegtbauke

project-development

Design and validate LLM batch processing pipelines with staged architecture and cost estimation.

Community
Advanced
SamyakJhaveriSamyakJhaveri

post-eval

Verify eval batch results and generate a concise summary.

Community
Advanced
curiositechcuriositech
2

dag-quality

Validate DAG node outputs against schemas and reliability criteria.

Official
Advanced
CarbeneAICarbeneAI
9

OllamaResearcher

Route sensitive research queries to a local Ollama API endpoint.

Community
Intermediate