How poco-ai/agentero works
Agentero is a local-first research workspace that brings paper discovery, library management, structured reading, notes, and AI-agent collaboration together. A typical workflow is to discover or import a paper, retain it in the vault, read and annotate its PDF, and build linked Markdown notes or ask an agent questions using research context. Major documented branches also include translation, citation tools, Zotero interoperability, RSS feeds, CLI automation, and remote or cloud synchronization. The graph uses README-documented capabilities where implementation excerpts were unavailable; arrows are limited to documented workflows and sampled call ownership.
Main entry points: engine-provider.tsx PDF reader · mod.rs Paper import · service.rs Paper catalog · App.tsx Desktop workspace
Components 21
Workspace
- Desktop workspace
src/App.tsx - Tauri host
src-tauri/src/app/handlers.rs - Headless CLI
cli/src/main.rs
Paper library
- Paper discovery
src-tauri/src/features/paper/discovery/mod.rs - Paper import
src-tauri/src/features/paper/import/mod.rs - Paper catalog
src-tauri/src/features/paper/catalog/service.rs - Paper analysis
src-tauri/src/features/paper/analyze/mod.rs
Reading and notes
- PDF reader
src/components/viewer/pdf/engine-provider.tsx - PDF features
crates/agentero-core/src/features/pdf/mod.rs - Markdown editor
src/components/editor/markdown-editor.tsx - Wiki notes
crates/agentero-core/src/features/markdown/wiki/mod.rs - Translation
crates/agentero-core/src/features/translate/mod.rs
Agent interfaces
- Agent panel
src/components/agent/agent-panel.tsx - Agent runtime
src-tauri/src/features/agent/runtime/mod.rs - MCP server
src-tauri/src/integration/mcp/server.rs
Interoperability
- Vault operations
crates/agentero-core/src/features/vault/mod.rs - Zotero integration
src-tauri/src/features/paper/zotero/mod.rs - Zotero Connector
src-tauri/src/integration/connector/server.rs - Vault sync
src-tauri/src/integration/sync/engine.rs
Other components
- Researcher
- S3 or WebDAV
Connections 22
- Researcher → Desktop workspace: uses
- Researcher → Headless CLI: automates
- Desktop workspace → Tauri host: invokes commands
- Researcher → Paper discovery: browses papers
- Paper discovery → Paper import: selects papers
- Researcher → Paper import: imports papers
- Paper import → Paper catalog: adds to library
- Paper catalog → Vault operations: stores locally
- Researcher → PDF reader: reads papers
- PDF reader → PDF features: uses PDF features
- PDF reader → Paper analysis: requests analysis
- Researcher → Markdown editor: writes notes
- Markdown editor → Wiki notes: resolves wikilinks
- PDF reader → Translation: requests translation
- Researcher → Agent panel: asks questions
- Agent panel → Agent runtime: sends interactions
- MCP server → Paper catalog: retrieves library context
- Researcher → MCP server: connects client
- Researcher → Zotero integration: migrates library
- Researcher → Zotero Connector: saves papers
- Researcher → Vault sync: starts sync
- Vault sync → S3 or WebDAV: transfers manifests and blobs