Repository intelligence

odysseus-dev/odysseus

An open-source, self-hosted personal workspace that merges AI chat assistants, multi-step deep research, and local model serving with daily productivity tools like Markdown editing, email, tasks, and CalDAV calendar sync.

CLOUDM0N decision

REVIEW BEFORE ADOPTION

Trust REVIEW · 10/100

Good fit if

Users wanting to consolidate their private productivity tools like email, calendar, and documents with local or API-driven AI agents in a single dashboard.

Watch out for

You prefer a fully managed cloud SaaS platform and do not want to manage local servers or Docker containers.

Practical intelligence

What matters before you adopt it

Decision first · detail on demand

Problem it solves

The friction of toggling between isolated applications for emails, calendars, writing documents, running search engines, and launching conversational local or API-based AI models.

Best for

Users wanting to consolidate their private productivity tools like email, calendar, and documents with local or API-driven AI agents in a single dashboard.

Developers seeking a self-hosted, AGPL-licensed environment to run web-researching agents and blind side-by-side model tests.

Main trade-offs

You prefer a fully managed cloud SaaS platform and do not want to manage local servers or Docker containers.

Why it stands out

Highly feature-complete, containing native IMAP/SMTP email, CalDAV sync, and a Markdown/HTML document editor.

Supports local or API-based models, MCP, shells, skills, and memory systems for custom agents.

Licensed under the copyleft AGPL-3.0-or-later standard ensuring open development.

Trust & CVEs

Security evidence without the noise

Trust remains a decision signal; CVEs and scanner evidence explain what is driving the risk.

Security findings

CLOUDM0N scanner findings

Critical

High

Medium

Low

View trust evidence & security findings▼

Why this score

No trust rationale was stored for this scan.

CLOUDM0N findings

HIGH

Remote download piped or chained into a shell requires manual review. 5 sample match(es) found.

HIGH

Container configuration requests host-level control or isolation bypass. 5 sample match(es) found.

MEDIUM

Dynamic code execution pattern detected. 5 sample match(es) found.

MEDIUM

The project can spawn operating-system processes; review command construction and input handling. 5 sample match(es) found.

MEDIUM

Broad permission or elevated-command pattern detected. 5 sample match(es) found.

MEDIUM

Dockerfile does not end with an explicit non-root USER.

Architecture from code22 modules · 49 edges▼

Structural evidence

Modules and dependency edges extracted from repository code. This is code evidence, not README inference.

Code files

1328

Modules

Dependency edges

Core modules

(root)

3 files

companion

3 files

core

core utilities

11 files

integrations

2 files

mcp_servers

5 files

routes

route handlers

86 files

scripts

18 files

services

40 files

src

109 files

src/agent_tools

11 files

src/model_capability_readers

10 files

src/search

8 files

Dependency flow

(root)→companion

(root)→core

(root)→routes

(root)→services

(root)→src

(root)→src/agent_tools

companion→core

companion→src

core→src

mcp_servers→core

mcp_servers→routes

mcp_servers→src

routes→core

routes→services

Detected languages

Python · JavaScript · Shell · Swift · TypeScript

Detected frameworks

FastAPI · NumPy · Pydantic · SQLAlchemy

Architecture evidence details

flowchart TD
    %% repo — high-level architecture (DRAFT, refine me)
    n0["(root) · 3 files"]
    n1["companion · 3 files"]
    n2["core · core utilities · 11 files"]
    n3["integrations · 2 files"]
    n4["mcp_servers · 5 files"]
    n5["routes · route handlers · 86 files"]
    n6["scripts · scripts · 18 files"]
    n7["services · services · 40 files"]
    n8["src · 109 files"]
    n9["src/agent_tools · 11 files"]
    n10["src/model_capability_readers · 10 files"]
    n11["src/search · 8 files"]
    n12["src/tools · tool implementations · 11 files"]
    n13["static · 169 files"]
    n14["tests · tests · 842 files"]
    n0 --> n1
    n0 --> n2
    n0 --> n5
    n0 --> n7
    n0 --> n8
    n0 --> n9
    n1 --> n2
    n1 --> n8
    n2 --> n8
    n4 --> n2
    n4 --> n5
    n4 --> n8
    n5 --> n2
    n5 --> n7
    n5 --> n8
    n5 --> n9
    n6 --> n2
    n6 --> n8
    n7 --> n2
    n7 --> n5
    n7 --> n8
    n7 --> n11
    n8 --> n2
    n8 --> n5
    n8 --> n7
    n8 --> n9
    n8 --> n11
    n8 --> n12
    n9 --> n2
    n9 --> n8
    n9 --> n11
    n11 --> n7
    n12 --> n2
    n12 --> n5
    n12 --> n7
    n12 --> n8
    n14 --> n0
    n14 --> n1
    n14 --> n2
    n14 --> n4
    n14 --> n5
    n14 --> n6
    n14 --> n7
    n14 --> n8
    n14 --> n9
    n14 --> n10
    n14 --> n11
    n14 --> n12
    n14 --> n13
    class n5,n7 api
    class n2 core
    class n6 infra
    class n12 shared
    class n14 test
    classDef api fill:#af7aa1,color:#ffffff,stroke:#8a5f7f
    classDef core fill:#e15759,color:#ffffff,stroke:#b04547
    classDef infra fill:#b35c00,color:#ffffff,stroke:#8f4a00
    classDef shared fill:#79706e,color:#ffffff,stroke:#5d5654
    classDef test fill:#499894,color:#ffffff,stroke:#397975

Evidence, security & integrations▼

Integrations

MCP (Model Context Protocol)IMAPSMTPCalDAVDocker

Security notes

• Keep 'AUTH_ENABLED=true' active for any deployment that is accessible over a network.

• Set 'LOCALHOST_BYPASS=false' outside of local development contexts.

• Do not expose raw model or service ports to the public internet.

• Keep private data out of Git version control.

Still unknown

• The README does not specify any operational trade-offs or constraints of using the software, leaving the trade_offs array empty.

• No information is provided regarding the specific database engine, file storage format, or persistent memory systems used locally.

• The text does not state the minimum RAM, storage space, or specific GPU model requirements necessary to run local model serving.

Adoption guidance▼

Adopt if

+ You want an integrated, offline-first personal workspace that combines document editing, email, calendars, and web search with robust AI model testing.

+ You need a self-hosted productivity system that respects data privacy and runs under a permissive AGPL copyleft license.

Avoid if

– You prefer a fully managed cloud SaaS platform and do not want to manage local servers or Docker containers.

How it works & getting started▼

How it works

1.The user deploys the container environment using Docker Compose.

2.The user checks the container logs to find the first generated admin password.

3.The user navigates to the local interface on port 7000 and logs in.

4.The user integrates preferred AI model providers, tools, or syncs email and CalDAV accounts.

5.The user schedules agent tasks, runs deep research, or organizes workflows natively in the dashboard.

Getting started

→ Choose between the curated 'main' branch or the bleeding-edge 'dev' branch.

→ Start the containers using Docker Compose.

→ Retrieve the initial administrative password by running 'docker compose logs odysseus' once the containers are healthy.

→ Open http://localhost:7000 in your browser to log in.

Agent handoff

Use with any agent

JSON API