Harbor vs ByteRover

ByteRover searches the files you upload. Harbor learns what your team decided.

ByteRover was a memory CLI for coding agents that kept what they saved as markdown files in a context tree. As read on 17 November 2026, it is document search: you upload files, and your agents get back the passage and its page. Asked whether it is a memory for your agent, its own FAQ answers "No. It is search." The memory CLI is archived on GitHub, and its last release was on 27 May 2026. Harbor is shared memory: it learns from where the team already decides, in Slack threads, PR reviews and docs, and serves the rules that apply to each session.

Free for you and one teammate. No card. No end date. Every line about ByteRover is from its own pages, read on 17 November 2026.

Why teams pick Harbor

Decide what gets in. See what gets used.

Three things Harbor does for the team behind the agents, each beside what ByteRover's own pages say about the same job.

You set the gate

Manual, Hybrid or Auto. Whatever you pick, secrets, install steps and dangerous commands wait for a person.

In ByteRoverNo review step is documented. Roles decide who can add files: an agent writes only in drives where its person is an Editor or Admin, and only if it was given permission to write.

You see what got used

Every rule shows how often it was served and how often an answer cited it.

In ByteRoverA Usage page shows what searches cost by person, agent or API key. Each passage names its file, page and revision.

You block a risky command

Block rules refuse it before it runs, in Claude Code, Cursor, Codex and Copilot.

In ByteRoverNot part of the product.

Side by side

Harbor and ByteRover, row by row.

HarborByteRover
What it isShared memory for a team's coding agents. It learns what your team decides and corrects, and serves the rules that apply to each session.Document search for agents, which it calls open-book search. You upload files to a drive, and agents and apps ask in plain language and get the passage and its page. It was a memory CLI for coding agents before; that repository is archived.
Where the knowledge comes fromSlack, GitHub, GitLab, Notion, Confluence, Jira, Linear, Google Drive and Granola, plus what agent sessions record.The files your team uploads: PDF, Word, PowerPoint, Excel, HTML, Markdown, text, email and code, up to 1 GB each. A connected agent saves a file only when you ask it to.
Who decides what gets inYou set the gate. Manual holds everything for a person, Hybrid adds routine rules and holds the important ones, Auto adds everything else. Secrets, install steps and dangerous commands wait in every mode.No review step is documented. Roles decide who can add files: an agent writes only in drives where its person is an Editor or Admin, and only if it was given permission to write.
How it reaches the agentHooks add the rules that apply to the session before the agent writes a line. The same rules are on an MCP server, the CLI and /harbor in Slack.When the agent calls its MCP server, or an app calls its REST API. A search takes about three seconds and returns up to 10 passages, each with its file, page and headings.
When a decision changesWhen a push deletes the file a rule is about, the rule is retired, or flagged if other repos still use it. A rule saved by an agent replaces one it reverses when its author could have retired the old one, with a Restore. Otherwise, and for anything read from a connected source, the pair waits for a person.You upload the file again. The new revision replaces the old one in search, and earlier revisions are kept until the file is deleted.
Seeing what was usedEvery rule shows how often it was served and how often an answer cited it.A Usage page shows what searches cost by person, agent or API key. Each passage names its file, page and revision.
Stopping a risky commandBlock rules refuse a command before it runs, in Claude Code, Cursor, Codex and GitHub Copilot, and hand the agent something to run instead.Not part of the product.
AgentsClaude Code, Cursor, Codex and GitHub Copilot (the CLI and VS Code).Any agent that supports MCP. Its site names Claude, ChatGPT, Cursor, Grok, Dot and Muse.
Open source and hostingThe CLI, harborloop, is MIT licensed on npm. The service is hosted.A hosted service. A licence and self-hosting for the search product are not documented. The archived memory CLI is under the Elastic License 2.0.
PriceSetup is self-serve and takes two minutes, with no call required. Free costs nothing for up to 2 people on one source, Pro is $24 per person a month billed yearly or $30 monthly, and Premium is $36, or $45 monthly.Pay per search, with no seats: $0.002 a search at medium effort, $0.006 at high, and $0.30 a GB a month past the first GB. A first team starts with a $5.00 balance and no card. Enterprise is custom.
Best forEngineering teams who want every coding session to start with what the team decided.Teams whose agents need the exact passage and page from documents they upload.

ByteRover's side comes from its own pages, read on 17 November 2026: byterover.ai · ByteRover docs · ByteRover pricing and billing · ByteRover agent tools · ByteRover agent permissions · ByteRover roles · ByteRover files and revisions · ByteRover CLI (archived). Out of date? Tell us.

Which one

Pick the one built for your job.

Pick Harbor if

  • You want the rules in the session before the agent thinks to search.
  • You want to see, rule by rule, which ones agents were served and which ones they used.
  • You want a rule retired when a push deletes the code it was about.
  • You want block rules that stop a destructive command before it runs.
Start free

Pick ByteRover if

  • You want agents to quote the exact passage and page from contracts, policies or specs.
  • Your knowledge is in PDFs and Office files more than in Slack threads and pull requests.
  • You want to pay per search, with no seats.
  • You need SOC 2 Type II today.
Can I run Harbor and ByteRover together?

Yes. They do different jobs. ByteRover looks things up in the documents you upload, and Harbor serves the decisions and corrections your team made to every session.

Can I start with Harbor today?

Yes. Sign up with Google or email, create your workspace and run npx harborloop init, with no card and no call. ByteRover can be signed up for today too.

Is Harbor open source?

The CLI, harborloop, is MIT licensed on npm. The service is hosted. ByteRover: A hosted service. A licence and self-hosting for the search product are not documented. The archived memory CLI is under the Elastic License 2.0.

Where do the claims about ByteRover come from?

From ByteRover's own public pages, read on 17 November 2026, which are listed under the table. Products move, so if a line is out of date, tell us and we will fix it.

Your team’s rules,
in every agent, checked on every PR.

Nothing to migrate and nothing to write. Every agent starts with what your team settles, and two weeks in you see which rules they cite.

Free for you and one teammate. No card. No end date.