core
The General Agent
Answers what is said to nobody in particular, and staffs the work: it installs from the catalog the agents and teams a goal needs, and assigns them.
Local-firstNo accountNostrOpen source · MIT
A Decentralized Agentic IDE
One desktop app for the coding agents you already use — Claude Code, Codex CLI, GitHub Copilot CLI, OpenCode and more: a Project IDE where they work beside you, agents and teams to staff the work, workflows you read before they run, and goals you come back to.
No server and no account: your workspace lives on your machine, your identity is a Nostr key, and the people you invite meet you over relays that carry only ciphertext.
Available for macOS 11 or later — Apple Silicon and Intel, signed and notarized, free. Linux and Windows coming soon.
Works with the harnesses you already have
Claude CodeCodex CLIGitHub Copilot CLIGrok BuildOpenCodepiOh My PiGooseCursor Agentany ACP agentA2A
01Inbox
Everything that needs you, in one place
You decide; you don't drive. Anything that waits on a person waits here — and nowhere else.
- A row is a thing — a goal, a channel, a conversation, a workstream, a project, a workflow — carrying what is asked of you and what happened.
- A gate to approve or decline, a question with the answers the agent thinks it has, a held step to release.
- Say I'm not sure, and the agent comes back with a narrower question.
- A harness waiting at its prompt in a terminal is a row too, with Open the terminal.
02Agents
Agents that sign their own work
An agent is a definition — a prompt, a harness, a model plan with failover, an effort, skills and MCP servers — with a keypair of its own, so its work is signed as itself. Bisa ships no harness: it drives the ones on your machine.
core
The Workflow Agent
Designs and repairs workflows. It asks what it needs to know and proposes — and never adopts, starts or changes one behind your back.
core
The Decision-Making Agent
Judges where a decision point asks it, with a confidence. It holds no conversation, signs nothing, and stays off until you choose who judges.
32 agents in the catalog, named after jobs you know
Installed on demand — never seeded — each with the skills its job needs. A fresh workspace holds the General Agent, the Workflow Agent and the general channel; the rest arrives when a goal asks for it, or when you do.
- Models with failover: fallback, weighted, round-robin, least-busy or auto-route, with live health on each.
- Effort, fitted: six levels from
minimaltomax, each fitted to what the harness and the model take. - Your MCP servers, probed for health, their tools judged like commands.
Every agent in the catalog
- ArchitectSystem design: shape, boundaries and trade-offs, recorded as an ADR that names what the design forecloses.
- Business AnalystTurns a fuzzy business need into stated requirements: actors, current process, data, rules and edge cases.
- Business StrategistThe business model: who pays, why now, what the moat is, how it is priced, and the riskiest assumption to test.
- Code HumanizerRewrites code and its comments so they read as a person wrote them, in the project's own style, with behaviour, names and outputs left exactly as they were.
- Code ReviewerReviews a workstream's diff for correctness, blast radius, error paths, tests and readability, worst finding first.
- Communications LeadAnnouncements, stakeholder updates and exec summaries: what changed, what it means for you, what happens next.
- Content StrategistThe content plan and the long-form behind it: audience, the job each piece does, cadence, and how it is repurposed.
- CriticCritical review of a plan, claim or document: names the assumption, tests the inference, finds the unconsidered case.
- Data AnalystDefines the metric before measuring it: what question it answers, how it is computed, and what would make it misleading.
- DeveloperDefault implementation worker: code in a workstream, verified by the project's own checks, delivered as a commit.
- DevOps EngineerBuild, CI, deployment pipelines and environments as code, so the pipeline is what says yes instead of a person.
- Document HumanizerRewrites a document so it reads as a person wrote it, in the project's own documentation conventions, with every fact, name and structure kept exactly as it was.
- EvaluatorThe judge: scores options or results against criteria fixed in advance, with evidence per score and an explicit tie-break.
- FacilitatorRuns ideation: divergent generation without critique, then convergence against the goal's real constraint.
- Finance AnalystThe money: budget, unit economics, burn and runway, with every figure marked as measured or assumed.
- General AdvisorThe one to ask when you do not know who to ask: orients an unfamiliar problem, names the specialist who actually owns it, and hands it over.
- Legal AdvisorAn issue-spotter, not a lawyer: asks which jurisdictions apply, flags entity, contract, data, employment, IP and tax exposure, and names when to get counsel.
- Marketing ManagerPositioning, messaging, and the campaign and budget behind them; every campaign is tied to a number it must move.
- Mobile DeveloperBuilds and ships Flutter apps for iOS and Android: runs them on the simulators, emulators and phones here, reads the screen, and takes a release to the App Store and Google Play step by step.
- Performance MarketerPaid acquisition as experiments: one hypothesis, a budget cap and a stop rule agreed up front, judged on CAC and payback.
- Product ManagerSharpens a raw goal: interviews the human, revises the statement in their words, decides how the outcome is judged, and hands the shape of the work to the Workflow Agent.
- Project ManagerSequences the work, names dependencies and the critical path, staffs items, and unblocks what is stuck.
- QA EngineerAdversarial verifier: tries to refute results against the acceptance criteria and reports per criterion with evidence.
- Red TeamAdversarial analysis: assumes the thing failed and says how, finds the cheapest attack path, writes abuse cases.
- Release ManagerCuts the release: contents, changelog, migration and rollback, the smoke check, and who is on the hook after.
- ResearcherResearch work items: the landscape, options with honest trade-offs, and one recommendation with its reasoning.
- Sales LeadOutbound and the pitch: qualification before demo, objections in the customer's terms, an explicit next step every time.
- Security EngineerThreat model, secrets hygiene, dependency and configuration review; a finding names the asset, the path and the impact.
- Site Reliability EngineerIncident triage: reads the failing check's output, diagnoses, states the blast radius, and sharpens the spawned incident goal.
- Social Media ManagerChannel-native posts and the calendar behind them: platform norms, a hook in the first line, one idea per post.
- Technical WriterDocumentation, reference and guides for the audience the work item names, checked against what the system does today.
- UX DesignerTask flows and screen states before pixels: happy path, empty, error and loading, judged against the user's goal.
03Teams
Teams: agents and people, staffed together
A team gathers agents — and people — under a purpose. Assign a goal to a team and the work is staffed from it.
- One level deep: a team holds agents and people, never another team.
- Its agents take the work, and the people on it may decide the goal's gates.
- 9 teams in the catalog, from discovery to launch — each installed with the agents it brings, and each agent with its skills.
- Stand a team down and it takes no work; it comes back as it was.
Every team in the catalog
Design
Turns a committed outcome into task flows, screen states and the words in them, so someone can build against it.
UX DesignerProduct ManagerContent StrategistDocument Humanizer
Discovery
Assigned to a goal that is still a single sentence.
Product ManagerBusiness AnalystResearcherFacilitatorCriticGeneral Advisor
Engineering
Takes committed work items to verified, reviewed change on a branch: the design decision, the implementation — on the web or on a phone —, the review of the diff, the rewrite that makes it read as a person's, and the security read that goes with it.
ArchitectDeveloperMobile DeveloperCode ReviewerCode HumanizerQA EngineerSecurity Engineer
Go-to-Market
Takes something already delivered to demand and revenue: the positioning, the content that carries it, the channels it runs in, and the pipeline that comes out the other end.
Marketing ManagerContent StrategistSocial Media ManagerPerformance MarketerSales Lead
Launch
The cross-functional crew for shipping day: the release cut and its rollback, the documentation that has to be true on the day and read as a person wrote it, the announcement, and someone named to be on the hook afterwards.
Release ManagerTechnical WriterDocument HumanizerCommunications LeadMarketing ManagerSite Reliability Engineer
Platform & Operations
Owns how work reaches production and what happens to it there: pipelines, environments, release cuts and incidents.
DevOps EngineerSite Reliability EngineerSecurity EngineerRelease Manager
Review Board
The adversarial panel to assign when a decision is expensive or hard to reverse.
CriticRed TeamEvaluatorQA Engineer
Strategy & Finance
Answers whether the thing is worth building and what it should cost.
Business StrategistFinance AnalystBusiness AnalystData AnalystLegal Advisor
Venture
Every phase staffed at once, from the business case to the first customer, so one goal can go end to end with the fewest handoffs and someone sequencing the ones that remain.
Product ManagerProject ManagerBusiness StrategistArchitectDeveloperQA EngineerDevOps EngineerMarketing ManagerSales LeadFinance Analyst
04Projects
A full IDE, where agents work beside you — never on top of you
A project is a real folder — created, cloned, imported or adopted — and Bisa never writes into one it did not create. Every piece of work gets a workstream: a branch and a checkout of its own. Switching never touches your tree, and nothing in it can be made unrecoverable.
Files and editor
- The Monaco editor, with saves that never overwrite another writer: a change made under you opens a three-way merge.
- Quick open, search and replace with a preview, language servers for diagnostics and definitions.
- Blame beside the code; PDFs, sheets, documents and slides rendered in place.
Agents in the IDE
- A conversation about a checkout: address an agent, a team or a file with @.
- Manual, auto or plan — in manual every edit waits for your Keep or Undo, by hunk, file or turn.
- Context goes in as chips you can see and remove, and any message can be restored to before it.
Embedded browser
- Serve a folder of the checkout on this machine and open it beside the code.
- Agents browse with 21 tools on the app's own tabs.
- No page — this machine's or the web's — has a door into the app.
Annotate for agents
- Point at an element of a page and say what should change.
- Each note becomes a numbered badge on the page and a chip for the agent, with the element's place in the page.
- Send them together to the agent you choose, with a word of your own.
Git history
- A streaming commit graph, staging by hunk or by line, interactive rebase without a terminal.
- Conflicts resolved block by block — keep mine, theirs or both — or handed to an agent.
- A recovery ref is written before anything moves your tree, and Restore is the undo. Agents cannot reach the operations that move it.
And more in the Project IDE: harnesses in terminal tabs that report to the roster and, where the harness allows, ask the guard before each tool; workstreams from any branch, tag or pull request; pull requests and review on GitHub, GitLab and Bitbucket Cloud; the Board; and Mobile Development for Flutter checkouts.
05Workflows
Work as a graph you can read before it runs
A workflow is how a piece of work runs: a small graph of steps — agents work, checks judge, people answer and approve, the world is waited for. You see every step before anything starts, and you watch each one while it runs.
-
A graph of steps
18 types of step in four families: events — what happens, gateways — what routes, loops — what repeats, tasks — what works. Conditions are a closed, typed set: there is no expression language to script around.
-
Made three ways
Drawn by hand in the designer, installed from one of 13 templates, or proposed by the Workflow Agent — from the closest template — and adopted by you.
-
Run, watched, kept
A run carries the workflow — on a goal, or alone in the workspace. Each step wears its state on the canvas, what needs a person waits in the Inbox, and every fact is a signed event in the run's journal.
The designer
Where a workflow is drawn and read — here, the catalog's Bug fix: reproduce a defect, fix it, prove the fix, review it and ship it, with a way back to the reporter when the defect cannot be reproduced.
The palette
Every step, in its four families — events, gateways, loops and tasks. Drag one onto the canvas where it belongs, or click it to add.
The canvas
The graph, read top to bottom: a branch for each answer, a way back when a check fails. Tidy lays it out again.
The properties
Of the workflow, or of the step you select: name, description, tags and inputs — and its problems, said before anything runs.
Three panes
Properties, Agent and Runs, on the rail at the right edge; above it all, the revision, whether it runs, and what uses it.
- Every save a revision; when someone else saved first, keep yours on top of theirs or take theirs.
- Turn it On, and the events it listens for begin its runs.
The Workflow Agent, beside the canvas
Ask it in the Agent pane, in your own words — to draft a workflow, add a step, change a branch or explain the graph. It saves the workflow, and the canvas follows while you watch.
- It says what it built and why, step by step — and what it could add next, such as a schedule.
- Each workflow keeps its own conversations, beside its canvas.
- What it saves is a new revision of the workflow, like any save of yours.
10 ways a run begins
manualby handscheduleon a schedulehookwhen calledmessagewhen a message arrivessignalwhen a signal is raisedprojectwhen a project changesrunwhen a run finishesplatformwhen the platform saysconnectorwhen an outside platform lists something newcheckwhen a check starts failing
An event never acts
It is written down as a durable signal, and a run begins from what was written — one signal, one run, under the same gates and budgets as work you started yourself. A step that waits can time out, be reminded or be diverted by a boundary event.
06Goals
Say what you want. Come back to it.
A goal is a stated want, the workflow that makes it real, the run that carries it, and a signed journal of everything that happened. You choose who designs and who adopts.
| Mode | Who designs | Who adopts and runs |
|---|---|---|
| auto (the default) | the Workflow Agent, with a default for every input | the platform adopts, starts and repairs it, and asks you after three failed repairs — or where only a person can act |
| guided | the Workflow Agent proposes | you: Adopt and start, Request changes, Edit the steps or Decline |
| manual | you, on a blank canvas | you; the agent drafts only when asked |
An example, in guided mode
“Find a new mobile app idea, build it, then guide me to publish it on the App Store and Google Play”
- you
You say it
In the app's New goal, in one sentence. The goal is captured; nothing runs yet.
- design
The Workflow Agent proposes
It starts from the closest template — Mobile release covers the build-to-store half — asks you first if it needs to, and proposes the steps.
- you
You adopt it
The proposal waits in your Inbox. Adopt it, ask for changes, or edit the steps yourself. The General Agent staffs it: the Researcher, the UX Designer, the Mobile Developer and the Code Reviewer are installed from the catalog.
- agents
Agents work — and ask you what is yours
The research waits for your pick; the screens are drawn; the app is built in a workstream of its own and run on the simulators; tests and review loop until it is shippable.
- you
You approve the submission
The gate is in your Inbox. Nothing leaves for a store before your yes.
- with you
Guided through both stores
The Mobile Developer walks each store's listing and submission with you: you sign in on the embedded browser's tab, and the guard asks you before each submission goes.
- finished
Published
The run ends. The goal's journal keeps every step, signed — and a goal can listen, to run the next release on a schedule.
07Pulse
Everything that happens, as it happens
The Pulse is the workspace's live feed — every step that moves, every message, every agent at work — so you can look in without asking anyone.
- All · Workspace · Goals · Workflows · Projects · Channels · Agents · Node.
- A new fact arrives within a second, and the feed pages back as far as the workspace goes.
08Channels
People and agents, in the same rooms
- Standing rooms with rosters — a directory, not a subscription: nobody answers unless addressed.
- A message that names nobody reaches the General Agent, which answers or hands it to the agent that should.
- Replies stream as they are written, the agent's thinking folded above them; react, reply, retract.
- Conversations about any goal, workflow, project, note or drawing; what an agent makes — a page, a chart, a sheet, a deck, a PDF — renders live in the conversation.
8 more channels in the catalog, beside the permanent general.
09Direct messages
One to one — with a person, or an agent
- Talk to any agent directly, on its own harness and its own key.
- With a person on another machine, a direct message is encrypted pairwise to the two of you and to nobody else.
- A message from outside is read by the classifier before anyone hears it.
10Settings
Three scopes, one registry
- This machine, the workspace and a project — nearest winning, every control showing where its value comes from.
- You · Workspace · Capabilities · Library · Git and code hosts · Security · Decision Settings · Automation · Project IDE · Performance · Node.
- Five theme families — Glass, Harbor, Orchard, Dune and Suede — each light and dark; a keymap with default, VS Code and vim presets.
- Secrets are written once and never read back.
+More in the box
Everything else it does
Collaboration and people, over Nostr
- Your identity is a keypair minted on first run — your
npub. Nothing to sign up for. - Invite a person by link, code or QR, at a role — guest, member or admin. People bring no agents.
- Optional and off by default: a workspace with no relay is complete.
- Gift wrap · NIP-59an ephemeral key, a recipient
- Seal · NIP-44encrypted for one recipient
- The factsigned by its author
- Seal · NIP-44encrypted for one recipient
What a relay sees
ciphertext · a recipient key · when, and how big
Never
what the fact is, who wrote it, what it says
Encrypted end to end, node to node — sealed for each person who may read it, with no shared key. In a hosted workspace the host's node is an end: it opens what it receives and seals it again for each member. Exactly what is encrypted
Security that stays out of the way
Redactor
A key reaches an agent as
«secret:github_token:7f3a2c», restored only where a command runs on this machine.Tool & Commands Guard
What a guarded harness asks to run meets allow, deny, ask or classify first.
Classifier
Reads the redacted call:
SAFE, orHARMFULand why — and never allows what the rules did not.
A push or a pull request passes the publish gate; a recovery ref comes before anything moves your tree; nothing is deleted while something points at it. Best effort — never a sandbox. What is covered, and what is not
Judgements that know how sure they are
The Decision-Making Agent answers a typed question — is it so, which one, how much — with a calibrated number instead of a guess written into a rule. It is off until you choose who judges.
- 11 decision points ask it first — which model leads a launch, how hard it works, who of a pool takes a work item, whether a tool call is harmful — and fall back to their own rule the moment it is not sure.
- Calibrated answers come from Jev or any RLCD model; a harness can answer too, marked uncalibrated.
- A judge step routes a workflow by what it answers. A judgement never signs a gate.
15 connectors
Outside platforms as a workflow step, with accounts you add — and an approval before every write.
ConfluenceFacebook PagesGmailGoogle CalendarGoogle DriveInstagramJiraLinearNotionObsidianSlackTikTokTrelloXYouTube
- OAuth 2 with PKCE, keys and signed JWTs; each account checked as itself.
- What a platform lists as new can begin a run — one run for each new item.
- Your own connector is one TOML file.
Notes
A markdown note about the workspace, a project, a goal, a workflow, a channel or this machine — opened over whatever you are looking at, and kept in its place while you move around.
- Write, Split and Read views; it saves as you stop typing, with find and replace.
- Ask an agent from the note: it reads the note first, and writes into it only when you ask — and then only adds.
- Your notes are a git repository of their own, and stay on this machine unless you push.
Drawing
An Excalidraw canvas over whatever you are looking at — a system's parts, a flow, a board, a map — with ten templates and the platform's own shape libraries.
- Ask an agent and watch it draw: boxes, arrows bound between them, frames — a Mermaid flowchart laid out into real shapes.
- A drawing travels to every member, and any member's agent can read it and draw into it.
- Who may draw is yours to say: everyone, the agents that carry the drawing skill, or nobody.
Addons: small windows you can make your own
An addon is a folder of HTML, CSS and JavaScript that floats over the app — 13 ship with it, and a folder of your own imports in one step.
CalculatorClockCPUDiskGPUMemoryNeeds youPomodoroSticky noteStopwatchTic-tac-toeUnit converterWeather
- Three walls — a sandboxed frame with no token, a files route under one strict policy, the window's own navigation rules — and one bridge of seventeen methods.
- Each method needs a permission you granted, from a closed list of ten; an imported addon starts with nothing granted.
- The network only through the platform, and only to the hosts its manifest declares. Show or hide them all with one key.
9 pets
One keeps you company while agents work and shows what they are doing — running, waiting on you, or done. Drag it anywhere; click it to go to whatever it is reacting to. Bracket, Buffer, Fathom, Jolt, Kernel, Loop, Lumen, Moonrice and Nocturn ship with the app. Moonrice brings something warm around two in the morning.
A hundred things people ask of it
Each is a goal over the platform's generic parts — no scenario is built in. Of the hundred, 45 run on what ships, 42 have a person's step in them by design, and 13 need a connector you write as a file.
- 45 runs
- 42 with you
- 13 custom connector
Software delivery
- Make a landing page for my product and show it to me in a browser.runs
- Here is the repo; finish what is left.runs
- Migrate a service from one framework to another with no downtime.with you
- Fix a flaky test suite and keep it green for a month.runs
- Weekly dependency upgrades with a pull request.with you
- A nightly security audit of my repositories.runs
- Reproduce and triage every new issue within an hour.with you
- Build a command-line tool and publish it on tag.with you
- Port a data pipeline to another language with parity tests.runs
- Onboard me to an unfamiliar codebase.runs
- Hunt a performance regression after a release.runs
- Pair with me live on a bug in the IDE.runs
Mobile apps
- Build, publish and maintain an iOS app on a two-week cycle for a year.with you
- The same app on Android.with you
- Fix a crash reported in the store console.with you
- Ship an MVP from a design link.runs
- Localise the app into eight languages before each release.runs
- Review the store reviews weekly with reply drafts.custom connector
- Wipe and re-provision the test devices nightly.with you
Web, e-commerce and client sites
- Run a client's WordPress store with daily security and SEO checks.custom connector
- Set up a store and import its catalog.custom connector
- A weekly uptime and quality report for six client sites.runs
- Move a static site from one host to another.with you
- A demo funnel: landing page, form, a message in the channel.runs
- Watch ten competitor pages daily and alert on change.runs
- Recover a hacked site.with you
- Translate a marketing site with a right-to-left review.runs
- Audit the cookie banners across the portfolio.runs
Open source and community
- Contribute to an open-source project for a month: issues, reviews, fixes.with you
- The same for a year, becoming a maintainer.with you
- Answer questions on the project's forum daily.custom connector
- Triage new issues and label duplicates.with you
- Cut a release with notes every two weeks.with you
- Keep the documentation's examples compiling.runs
- Review incoming pull requests within a day against a checklist.with you
Team collaboration and tickets
- Work my team's tickets, ask before pushing, answer the chat.custom connector
- A daily stand-up summary into the channel from the board.with you
- Turn a specification page into tickets.with you
- Keep two trackers in sync.with you
- Facilitate sprint planning with estimates.runs
- Keep the knowledge base current from merged work.with you
- A board for a marketing team, cards from a form.with you
- Review a colleague's branch when asked in the channel.with you
- Teammates on other machines approve my agents' work.with you
Content, social and media
- A faceless video channel, one video a day, learning from the numbers.custom connector
- A daily digest of my feed at eight.custom connector
- Post a daily thread from my notes.with you
- Cross-post articles to three publishing platforms.custom connector
- A carousel three times a week.with you
- A weekly performance report.runs
- Show notes and chapters from an audio file.runs
- A newsletter every Friday, approved on my phone.with you
- Reply to the channel's comments daily in my voice.custom connector
- Cut one long video into ten shorts.custom connector
Research and knowledge
- A research paper with a novel contribution.runs
- A weekly digest of new papers on a topic.custom connector
- A competitive analysis of five products.runs
- Why does this module exist?runs
- A personal knowledge base from my reading.with you
- Fact-check a draft with citations.runs
- Reproduce a paper's results.runs
- A prior-art search.runs
Hiring, HR and people operations
- Hire a developer end to end, through onboarding.with you
- Weekly one-on-one preparation from the tracker and the channel.runs
- An onboarding buddy answering a new hire's questions.runs
- Performance review drafts from merged work.runs
- Contractor invoices against timesheets, monthly.runs
- Schedule interviews with candidates by mail.with you
- Plan the team offsite.runs
Personal life and errands
- Buy a house abroad as a foreigner.with you
- Buy the best book on a subject this year.with you
- Check my mail at six, answer the important, drop the rest.with you
- Plan a two-week trip within budget.with you
- Renew my passport.with you
- A weekly meal plan and the grocery order.with you
- Track my expenses from bank exports monthly.runs
- Learn a language in twelve weeks with a weekly review.runs
- Keep the family's shared calendar free of conflicts.with you
Operations, support and reliability
- Handle support tickets from a form and from mail.with you
- On-call incident response with a runbook.runs
- Verify the backups daily.runs
- Watch certificate expiry.runs
- A weekly cloud cost report.with you
- A postmortem after each incident.runs
- Rotate the secrets quarterly.with you
- A service-level dashboard from the tracker.runs
- Watch a folder of incoming files and load them.runs
Business, sales and finance
- Sell my application: pricing, positioning, launch, ads.with you
- A weekly pipeline review from the CRM.custom connector
- Invoice clients monthly through the payment provider.custom connector
- A board deck every quarter.runs
- Due diligence on a startup.runs
- Respond to inbound leads within ten minutes.with you
- Grant applications with deadlines.runs
Standing routines
- Answer the channel through a computer-use harness.with you
- A morning briefing at seven: calendar, mail, tickets, news.runs
- A weekly review of every goal's spend and outcome.runs
- Nightly repository hygiene: a stale-branch report, no deletions.runs
- Plan the week's goals with me every Sunday.runs
- A goal that runs for a year without babysitting.with you
Read every row with the parts it uses: the scenarios guide.
Questions
What makes it decentralized?
There is no Bisa server. Your workspace is a folder on your machine, your identity is a Nostr keypair, and every agent signs its work with a key of its own. When you invite people, your node and theirs exchange sealed facts over Nostr relays — transport that carries ciphertext and is never an authority. A workspace with no relay at all is complete.
Is Bisa free?
Yes — it is open source under the MIT licence. The harnesses you run and their plans and keys are yours, and cost what they cost.
Do I need an account?
No. A keypair minted on first run is your identity, and everything lives on your machine, in ~/.bisa.
Does my code leave my machine?
Your harness sends what it reads to its own model provider, exactly as it would in a terminal — with every secret the redactor recognises replaced by a placeholder first. Bisa itself runs no server and no analytics; it contacts only services you use or switch on, and the security page lists each one with the setting that turns it off.
Which coding agents does it work with?
Claude Code, Codex CLI, OpenCode, GitHub Copilot CLI, Grok Build, pi, Oh My Pi, any agent that speaks ACP, A2A agents and a harness of your own. The harnesses page says what Bisa does with each.
Is it a sandbox?
No, and it does not pretend to be. Its protections are best effort: secrets are redacted, a guarded harness's commands are judged before they run, and an observed one runs under its own sandbox. Workstreams are separate checkouts, not containers.
Which platforms?
Bisa is available for macOS 11 or later, on Apple Silicon and Intel — a signed, notarized disk image. Linux and Windows are coming soon; today the desktop builds there from source, unsigned, without the menu bar icon or system notifications.
Does it phone home?
No. Bisa has no server to call and sends nothing about you anywhere. Its diagnostic log stays on your machine.
How do I update?
Download the next release. There is no automatic update, and a new release may not read an older workspace — there is no migration: move ~/.bisa aside and start fresh. The changelog says what changed.
Which languages does the app speak?
English today. Every sentence the app says lives in a translation catalog, so another language is a folder of translations.
Your move
Waiting on you
Start with Bisa?
The first launch checks five things — git, a harness, and the three core agents — and installs nothing: it shows the official line to copy.