alexsosn/Agora

**Agora is a plugin marketplace that brings scholarly corpora, lexica, textual databases, and research services into AI-assisted philological workflows.**
It lets Claude Code and ChatGPT/Codex use research tools for Biblical Studies, Classics, Syriac, Ugaritic, Hittite, and related fields without requiring every resource to use the same backend or corpus format.
Agora currently includes four plugin families:
- **Context-Fabric** — structured local querying of registered Text-Fabric corpora, including BHSA, Ugaritic and Hittite corpora, and large Greek collections.
- **Perseus** — live Perseus/Scaife text discovery, passage retrieval, and full-text search.
- **Sefaria** — Jewish texts, translations, links and commentaries, dictionaries, topics, and manuscript resources through the official Sefaria MCP.
- **SEDRA** — Syriac word-form and lexeme lookup against SEDRA IV.
Agora plugins can also include **scholarly skills**: source-specific guidance that tells the agent how to interpret corpus features, avoid common mistakes, and produce more reproducible research queries.
New to Agora? Start with the Researcher user guide to choose a plugin, install it, check compatibility, find current research examples, or troubleshoot a problem.
What you can do
With the current plugins you can, for example:
- query morphology and syntax in the **BHSA Hebrew Bible**;
- work with **Ugaritic** and **Hittite** Text-Fabric corpora through the same Context-Fabric plugin;
- discover and load individual works from large **Greek Text-Fabric collections** without installing one plugin per text;
- retrieve passages and search Classical texts through **Perseus/Scaife**;
- fetch source texts, translations, linked commentaries, and other resources from **Sefaria**;
- look up **Syriac** word forms and lexemes in SEDRA;
- ask the agent to inspect corpus schema and annotation before constructing a query instead of guessing feature names;
- keep corpus-specific research guidance alongside the tools that use it.
The current Context-Fabric catalog contains **39 registered upstream resources**. Corpora are acquired lazily, so installing the plugin does not download all of them.
Context-Fabric also accepts user-supplied native TF corpora and feature modules through install_local_corpus, without adding them to the canonical catalog.
Installation
Agora currently targets **Claude Code** and **ChatGPT/Codex**. You only need to install the plugins relevant to your research.
Claude Code
Inside Claude Code:
You can install only one or two of these if that is all you need.
ChatGPT / Codex
If your workspace supports GitHub marketplace import:
1. Open **Workspace settings → Plugins**.
2. Choose **Add → Import marketplace**.
3. Use https://github.com/alexsosn/Agora as the source.
4. Leave the path empty; Agora's marketplace file is at the repository root.
5. Import the marketplace and enable the plugins you need.
Agora plugins declare MCP servers. Current GitHub-imported MCP plugins can therefore be labeled **Desktop only**; a Desktop-only plugin cannot run in ChatGPT web. Availability also depends on the current workspace, role, plan, and product surface.
For local Codex development/testing: