Tool Catalogue¶

Canonical documentation for every AI tool, framework, provider, agent, and infrastructure component used in or evaluated for this stack. One page per tool — no duplicates.

Current Sections¶

Directory	Category label	What lives here
ai_knowledge/	AI & Knowledge	LLM interfaces, knowledge bases, creative AI, local model frontends
agents/	Agents	Autonomous coding and task agents (Jules, Claude Code, Goose, Superpowers, etc.)
automation_orchestration/	Automation & Orchestration	MCP servers (FastMCP 3.1), workflow connectors, browser automation
benchmarking/	Benchmarking	LLM evaluation suites, leaderboards, multi-language coding harnesses (Inspect AI, MultiPL-E, etc.)
development_ops/	Development & Ops	AI coding assistants, DevOps tooling, documentation tools (OpenCode, MkDocs, etc.)
frameworks/	Frameworks	LLM application frameworks (LangChain, LlamaIndex, Dify, AutoGen, etc.)
infrastructure/	Infrastructure	Inference engines, vector stores, proxies, fine-tuning suites (vLLM, SGLang, Qdrant, PEFT, TensorRT-LLM, Triton)
providers/	Providers	API and cloud AI providers (OpenAI, Anthropic, Google AI Studio, Azure AI Search, DeepSeek, etc.)
calendar_tasks/	Calendar & Tasks	Time management, scheduling, CalDAV integrations
intake_storage/	Intake & Storage	Standard protocols and storage tools (CalDAV, MinIO, S3, etc.)
process_understanding/	Process & Understanding	Document analysis, OCR, web crawling, observability (Crawl4AI, Docling, Prometheus, Tempo, Logfire, Grafana)
orchestration/	Orchestration	Workflow engines and data pipeline tools (Temporal, Airflow, Dagster, Kestra, etc.)
enterprise/	Enterprise AI	Workplace-specific productivity suites, identity, and search (Glean, Hebbia, Okta, Proton Mail, etc.)

Note on taxonomy drift: automation_orchestration/ and orchestration/ currently coexist. Per standards.md, orchestration/ is the canonical location for workflow DAG/pipeline engines. New orchestration tool pages go there; automation_orchestration/ content focuses on MCP protocols and agent integration harnesses.

Self-hosted services (Paperless-ngx, n8n, Nextcloud, Ollama, etc.) live in Services — they carry operational deployment context alongside tool documentation.

How to Find the Right Tool¶

I need to automate a multi-step task → Automation & Orchestration, Orchestration, or Playbooks

I need to run a model locally → Infrastructure or Services → Ollama

I need an LLM-powered app framework → Frameworks

I need to pick an API provider → Providers and Knowledge Base → Model Routing Guide

I need an autonomous agent to write code → Agents

I need to evaluate or compare models → Benchmarking and Knowledge Base → Model Comparison

I need to ingest or process documents → Process & Understanding or Services → Paperless-ngx

Catalogue Rules¶

Every page must follow the standard template with these required sections:

## What it is · ## What problem it solves · ## Where it fits in the stack · ## Typical use cases · ## Strengths · ## Limitations · ## When to use it · ## When not to use it · ## Related tools / concepts · ## Sources / references

AI-authored pages must also carry:

## Sources / References
- [Official docs](https://example.com)

## Contribution Metadata
- Last reviewed: YYYY-MM-DD
- Confidence: high | medium | low

See Standards for the full taxonomy and dedup policy, and Contributing for how to add a new tool page.

Sources / References¶

Contribution Metadata¶

Last reviewed: 2027-01-07
Confidence: high