Sesori Mobile Coding Cockpit

Keep computer-based AI coding sessions within reach from your phone.

Stars
★ 143
Last updated
1d ago
License
NOASSERTION
Primary language
Dart

At a glance

How it runs
Desktop appCLI
Works with
Universal · cross-platformCodex · Claude Code
Setup effort
Medium · a few setup steps
You'll need
Sesori mobile appAI coding assistant CLIShell / CLINetwork accessLocal filesystem
Typical use
A developer stepping away from their desk who needs to answer a Codex or Claude Code session prompt from a phone.
Not a fit if
  • Teams that need phone-to-machine control to work fully offline
  • Users who need to log in to assistant providers from their phone
Website
sesori.com ↗ set on GitHub
Maintenance
Commits in the last 30 dayslast commit 2026-11-20latest release v1.15.0 (2026-11-20)
Source review
67/100 · Some gaps

What does this agent do, and when should you use it?

Sesori combines a Bridge CLI that runs on your computer, iOS and Android clients, and shared encryption and protocol components. The Bridge connects locally to assistants such as OpenCode, OpenAI Codex CLI, GitHub Copilot CLI, and Claude Code, then communicates with the phone app through an encrypted WebSocket relay. From the phone, users can browse projects and sessions, answer assistant questions, approve steps, stop or restart tasks, and review diffs. Application data, prompts, and replies are encrypted end to end between phone and computer; the relay routes encrypted traffic. The Bridge supports macOS, Linux, and Windows, while the mobile app requires iOS 15 or later or Android 8.0 or later.

Start sesori-bridge on a computer and it connects to local AI coding assistants, pairing with the phone app through an encrypted relay. The phone client lists projects and active sessions, accepts typed or voice input, lets users answer questions, approve steps, stop or restart tasks, and read diffs. The README lists OpenCode, OpenAI Codex CLI, GitHub Copilot CLI, Google Antigravity, Cursor, Claude Code, Hermes Agent, Pi, Oh My Pi, DeepSeek Harness, and Grok Build; integration methods vary by assistant. Devices perform an ephemeral X25519 key exchange and encrypt messages with XChaCha20-Poly1305.

  1. A developer stepping away from their desk who needs to answer a Codex or Claude Code session prompt from a phone.
  2. An engineer running an assistant on a remote machine or VM who wants to check progress and approve steps on mobile.
  3. A developer running several supported coding assistants who wants to browse their sessions in one phone app.
  4. An individual developer running a long coding task who wants a notification when it finishes or needs input.
  5. A reviewer who wants to inspect diffs from an AI coding session on a phone.

Key setup facts

Pulled from the install and usage notes below. The project README has the current values.

Install / run
  • curl -fsSL https://sesori.com/install.sh | bash
  • irm https://sesori.com/install.ps1 | iex
  • npx @sesori/bridge
  • bunx @sesori/bridge
Install notes cover
Windows · macOS · Linux · Android

How do you install or deploy this agent?

Install the Sesori mobile app (iOS 15+ or Android 8.0+), then install the Bridge on your computer. On macOS/Linux, run:

curl -fsSL https://sesori.com/install.sh | bash

In Windows PowerShell, run:

irm https://sesori.com/install.ps1 | iex

You can also bootstrap the Bridge with npx or bunx:

npx @sesori/bridge
bunx @sesori/bridge

Sign in with the same Sesori account on the phone and computer. The provided material does not state the price of the app or Bridge.

How do you use this agent?

Start the Bridge in a terminal on your computer:

sesori-bridge

Sign in to the phone app with the same Sesori account, choose an assistant, and start a session. Provider authentication happens locally; for example, Copilot CLI supports copilot login, while Pi uses its /login command. Use the phone to continue interacting with sessions, review diffs, and receive notifications.

What are this agent's strengths and limitations?

Pros
  • Supports multiple coding assistants, with several integrations usable at once on a single Bridge.
  • The phone client can answer questions, approve steps, stop or restart tasks, and review diffs.
  • The Bridge and client protocol are source-available; the README describes X25519 key exchange and XChaCha20-Poly1305 encryption.
  • The Bridge runs on macOS, Linux, and Windows and can be used on headless remote machines.
Limitations
  • Requires installing both the Bridge and mobile app and signing into the same Sesori account on each.
  • Remote interaction uses a relay, so it requires network connectivity.
  • Assistant capabilities differ; for example, Copilot CLI does not currently forward ask_user interactions over ACP.
  • Provider credentials and login remain on the Bridge machine; the README does not document phone-based provider login as a general feature.
  • The repository metadata says NOASSERTION while the README describes an FSL-1.1-ALv2 source-available license, so it should not be assumed to use a conventional open-source license.

How does this agent compare with similar options?

Compared with using OpenCode, Codex CLI, or Claude Code only in a terminal, Sesori adds mobile session browsing, remote interaction, diff review, and notifications; it still depends on the relevant assistant being installed and authenticated locally.

Key facts side by side with the most closely related agents.

Agent Source review Form / cost Stars Updated Language Full support on
Sesori Mobile Coding Cockpit This agent 67 · Some gaps Desktop app ★ 143 1d ago Dart Codex · Claude Code
HAPI Remote Agent Console 69 · Some gaps CLIFree + model costs ★ 5.4k 1d ago TypeScript Codex · Claude Code
DSH Go 78 · Good Desktop appFree + model costs ★ 29k today Dart —
Happier 62 · Some gaps Desktop appFree + model costs ★ 2.2k today TypeScript Codex · Claude Code

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Some gaps
67/ 100 5-point scale 3.4 / 5
This review covers the source as of Nov 12, 2026. The project has shipped 5 stable release(s) since, which have not been re-reviewed. The score and verdict describe the reviewed version only. Recent releases ↓
Trust 19/29
Reliability 8/14
Adaptability 15/18
Convention 12/18
Effectiveness 7/13
Verifiability 6/8
Why each dimension lost points
Trust19 / 29 · 3.3/5

SECURITY.md honestly states the relay is trusted for key delivery and not hardened against a malicious relay, and discloses compression side channels, push previews leaving the E2E channel, and feedback storage — data_flow_transparency earns 3. However Pi launches with --approve by default, OMP defaults to yolo, and the YOLO auto-approval setting removes the human safety gate, so least_privilege and user_confirmation are capped at 2. Dependency security is weak: .mcp.json uses npx -y without full hash pinning, and installers use curl|bash / irm|iex with only SHA-256 verification, so 1. Rollback is documented for bridge updates with in-place rollback and startup reconciliation, 2. Source attribution has LICENSE and copyright but publisher is unverified, 2.

Reliability8 / 14 · 2.9/5

Documentation is internally consistent on capability boundaries (Copilot no ask_user, Grok delete only closes session), self_consistency 2. Dependency availability relies heavily on external CLIs (Grok, Hermes, Antigravity) and PATH probing, with Antigravity real OAuth and authenticated sessions unverified, so 1. Failure messages in release_scheduler and regression docs show structured errors and redacted logs, 2.

Adaptability15 / 18 · 4.2/5

README and AGENTS.md clearly define multi-harness, multi-surface scenarios, audience_and_scenarios 3. capability_boundaries are documented per harness in <details> blocks and HARNESS_CAPABILITIES references, 3. trigger_precision is weaker: skill descriptions are broad (add-analytics covers many scenarios) and AGENTS.md mandates plan-maker which is not included in the packet, 2. environment_fit covers macOS/Linux/Windows and mobile, but some platform evidence is missing (Intel Mac, Linux/Windows native unverified), 2.

Convention12 / 18 · 3.3/5

Information architecture is clear (README, docs, regression, skills layered), 3. Install notes give 3 steps and multi-platform commands but curl|bash and npx bootstrap lack verification detail, 2. Naming stability has version sync tooling and compatibility comment conventions but no changelog, 2. Examples and FAQ exist as <details> notes but no standalone FAQ, 2. Known limitations are thorough across harness notes and regression Known Limitations, 3. License is FSL-1.1-ALv2, non-OSI with competing-use restriction, 2. Versioning/changelog only shows release badge and version sync, no CHANGELOG file, 1. Maintenance responsibility is unclear with unverified publisher, only email and Discord, 1.

Effectiveness7 / 13 · 2.7/5

Output usability is concrete for phone-based session control, 2. Marginal value is unified cross-harness control with E2E encryption, but competitors exist, 2. Cost-benefit requires installing Bridge, pairing accounts, relying on external CLIs and relay service, plus YOLO default risk, 1.

Verifiability6 / 8 · 3.8/5

claim_traceability: README claims link to docs and regression documents, 2. cross_source_corroboration: SECURITY.md and README agree on E2E, regression docs and README agree on harness limits, 2. fact_inference_separation: docs explicitly separate verified from unverified (e.g., Antigravity real OAuth unverified), 3.

Risks and how to mitigate them
  • Pi launches with --approve by default and OMP defaults to yolo; YOLO auto-approval removes the human safety gate — use only on trusted projects.
  • Installers use curl|bash and irm|iex with only SHA-256 verification, not OpenPGP signatures, creating supply-chain risk.
  • .mcp.json uses npx -y without full hash pinning; pin versions and hashes.
  • The relay is trusted for key delivery; a malicious relay could MITM, and docs explicitly do not harden against it.
  • Push notification previews leave the E2E channel and may pass through Apple/Google push services.
  • Publisher identity is unverified; maintenance responsibility and update path are unclear.
  • License is FSL-1.1-ALv2 with competing-use restrictions, not standard open source.
Evidence confidence: Low Reviewed Nov 12, 2026 Reviewed revision 8c26051e83a0 New commits since this review; the score may not cover them
See the full review method →
Maintain this project? Add the FARS badge to your README

The badge shows the current review score and verdict and links back to this page. It refreshes about once an hour after a re-review.

FARS

[![FARS review](https://followagents.com/badge/sesori-apps-monorepo.svg)](https://followagents.com/en/agents/sesori-apps-monorepo)
<a href="https://followagents.com/en/agents/sesori-apps-monorepo"><img src="https://followagents.com/badge/sesori-apps-monorepo.svg" alt="FARS review"></a>

Recent releases

Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.

  1. v1.16.0-internal.1254 Internal v1.16.0-internal.1254 Pre-release
    Release notes excerpt
    What's Changed App Changed - 🌱 Bump version to 1.16.0 (#2856) - 🌿 Show one scrollbar for a prompt scrolling inside the new task and new session pages (#2857) - 🌿 Keep the image viewer's close button clear of the macOS traffic lights (#2859) Bridge Changed - 🌱 Bump version to 1.16.0 (#2856) All PRs merged * 🌱 Bump version to 1.16.0 by @AlexDochioiu in https://github.com/sesori-ai/sesori_apps_monorepo/pull/2856 …
    View on GitHub ↗
  2. v1.15.0 v1.15.0 (build 1252)
    Release notes excerpt
    ✨ Highlights - 🤝 Agents can now use Sesori's own tools. Claude Code, Codex, Pi, Oh My Pi, Grok and GitHub Copilot sessions can read your scheduled tasks and their runs, and propose new tasks or changes that you approve on your phone. - 📝 Question pages from your agents. An agent can send you a page of questions with images and answer options, and you fill it in from the app instead of typing replies in chat. …
    View on GitHub ↗
  3. v1.14.1 v1.14.1 (build 1229)
    Release notes excerpt
    What's Changed App Changed - 🌿 Open a handoff's full summary in a sheet instead of expanding the card (#2760) - 🌱 Give markdown headings in session summaries the app's type (#2761) - 🌱 Bump version to 1.15.0 (#2762) - 🌱 Let a new session take its launching row's place below pinned sessions (#2763) - 🌱 Show a created session under a chip or search while its launching row waits on All (#2764) - 🌿 Keep a waiting launch's session out of Running, Unread and search too (#2765) …
    View on GitHub ↗

FAQ

Is Sesori free?
The supplied material does not state the price of the Sesori app or Bridge. Assistant service eligibility or account requirements may apply separately; for example, Copilot CLI requires an eligible account.
Can the phone access my code and prompts?
The README says application data is encrypted end to end between phone and computer and the relay routes encrypted traffic. Push notification previews may include a short event snippet.
Can I control a computer session remotely without internet?
Offline remote control is not documented. Both Bridge and phone connect to a relay, which the README describes as using WebSockets.
Can I sign in to every assistant provider from my phone?
No general phone-based provider login is documented. The README says provider logins are generally local; its Google Antigravity notes separately describe personal Google OAuth from a current phone or desktop client.
Do all assistants support the same operations?
No. Protocols and capabilities vary by integration. For instance, Copilot CLI does not forward ask_user interactions, and the Grok Build ACP integration currently does not support image prompt attachments.
Does Sesori Mobile Coding Cockpit work on Windows?
Yes. The install notes on this page include a Windows command: irm https://sesori.com/install.ps1 | iex.
Is Sesori Mobile Coding Cockpit open source?
The source code is public on GitHub (sesori-ai/sesori_apps_monorepo), but the repository does not declare a recognised licence, so the rights to reuse or modify it are unclear. Ask the author or check the README before building on it.
Is Sesori Mobile Coding Cockpit safe to use?
FollowAgents' FARS review (Nov 12, 2026) scored Sesori Mobile Coding Cockpit 19/29 on trust and safety. None of the eight safety controls the review checks for was missing outright. The repository has changed since that review.
Is Sesori Mobile Coding Cockpit still maintained?
Yes. The latest commit on GitHub is from 2026-11-20, within the last 30 days. The latest release is v1.15.0 (2026-11-20).
Who makes Sesori Mobile Coding Cockpit?
Sesori Mobile Coding Cockpit is published on GitHub by the account sesori-ai. FollowAgents has not verified which person or company is behind that account.
What is Sesori Mobile Coding Cockpit's official website?
The project's GitHub repository lists https://sesori.com as its website. That link is set by the repository owner; the source code itself is on GitHub (sesori-ai/sesori_apps_monorepo).
Was this page helpful?
View on GitHub ↗ Install ↓

Compare agents like this one

The same FARS review applied across the shortlist this agent qualifies for.

Related concepts