DeepSource MCP server

What the model can do

DeepSource (code quality and security analysis): the accounts and repositories the token reaches, analysis runs with the issues each introduced, resolved and suppressed, issues on the default branch by category and severity, pull requests with their analysis summary, dependency vulnerabilities with severity, reachability and fixed versions, and dependencies that break the license policy; can suppress or unsuppress an issue for the whole team, switch analysis on or off for a repository, change its default branch, and set its quality gates (issue categories, priorities, metrics and thresholds).

Reads

• deepsource_list_accountsThe DeepSource accounts (teams and individual accounts) this token reaches, with the login and VCS provider every other tool needs.
• deepsource_list_repositoriesRepositories in a DeepSource account: whether analysis is on, private or public, default branch and latest commit analysed.
• deepsource_list_analysis_runsAnalysis runs on a DeepSource repository: status, branch, commit, when, and how many issues each introduced, resolved and suppressed.
• deepsource_get_runOne DeepSource analysis run by its run uid or by commit SHA: status, branch, commit and base, when, issues introduced, resolved and suppressed…
• deepsource_list_issuesIssues DeepSource raises on a repository's default branch: shortcode, title, category, severity and a short description…
• deepsource_list_pull_requestsPull requests DeepSource analysed on a repository, or one by number: title, state, branches, link, issues raised, resolved and suppressed…
• deepsource_list_vulnerabilitiesVulnerable dependencies DeepSource found on a repository's default branch: the CVE or advisory, severity, CVSS and EPSS scores…
• deepsource_list_dependenciesPackages used on a repository's default branch with their licenses and whether each breaks the team's license policy.

Acts, when you ask

These change something in DeepSource. The model is told to use them only when you ask for exactly that, and each one is granted by name: per team on a shared connection, by the owner on a personal one.

• deepsource_suppress_issueSuppresses an issue (by shortcode, e.g. PYL-E0712) for the whole DeepSource team, so no new analysis run in any of its repositories reports it…
• deepsource_unsuppress_issueRemoves a team-wide suppression of an issue in DeepSource, so new analysis runs report it again in every repository.
• deepsource_activate_repositorySwitches DeepSource analysis on for a repository, which starts analysing it and checking its pull requests.
• deepsource_deactivate_repositorySwitches DeepSource analysis off for a repository: commits and pull requests stop being analysed and checked until it is switched on again.
• deepsource_set_default_branchChanges the branch DeepSource treats as a repository's baseline; if analysis is on, it starts a new analysis.
• deepsource_set_category_gateSets whether a DeepSource issue category (security, bug risk, style…) is reported on a repository and whether it fails the pull request check…
• deepsource_set_priority_gateSets whether DeepSource issues of a priority (low, medium, high) are reported on a repository and whether they fail the pull request check…
• deepsource_set_metric_gateSets whether a DeepSource metric (line coverage LCV, branch coverage BCV…
• deepsource_set_metric_thresholdSets the threshold a DeepSource metric must meet on a repository (e.g. line coverage at least 80), for one language key or AGGREGATE…

Connect it

First connect DeepSource on Dashboard → Connections, then add Memtro to your client once. The DeepSource tools appear alongside the tools of everything else you have connected; nothing is configured per connector.

claude plugin marketplace add memtro-org/claude-memtro
claude plugin install memtro@memtro --config api_key=sk-memtro-...

{
  "mcpServers": {
    "memtro": {
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

{
  "mcp": {
    "memtro": {
      "type": "remote",
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

Who may use it

The key identifies the person, and the server offers that person the tools they may use everywhere else in Memtro. A shared DeepSource connection is used by the teams an administrator granted it to, tool by tool, on the Permissions page; a personal connection is used by its owner alone, and its tools that act stay off until the owner allows them. See Teams and permissions and Connectors.

Ask it like this

• › Which issues did the last analysis of our API repo introduce?
• › Which reachable vulnerabilities does the web repository have?
• › Which dependencies break our license policy?
• › How many issues did pull request 412 raise?
• › Set line coverage on the api repository to at least 80.

Questions

Is this an official DeepSource MCP server?

No. It is Memtro's MCP server with Memtro's DeepSource connector behind it: the connector reads DeepSource through its API with the credentials you connect on Dashboard → Connections, and DeepSource publishes nothing here. The same endpoint carries every other system you connect and Memtro's memory tools.

What does it cost?

Nothing on top of Memtro. One price per organisation by its number of users, from $50 a month for up to 10 users after a 7-day free trial, with every connector, the MCP server, jobs and memory included. You bring your own model keys.

Can it change anything in DeepSource?

Only when you ask, and only what it has been granted: suppress an issue for every repository in the team; unsuppress an issue for the team; switch analysis on for a repository; switch analysis off for a repository; change a repository's default branch (starts a new analysis); set an issue category's quality gate on a repository; set an issue priority's quality gate on a repository; set a metric's quality gate on a repository; set a metric's threshold on a repository. Every other tool reads, searches or lists. An administrator grants each acting tool by name per team, and a personal connection acts only after its owner allows it.

Which clients work?

Any MCP client that speaks Streamable HTTP: Claude Code, Cursor, Claude Desktop, OpenCode, Codex, Zed, Windsurf, Cline and the Anthropic and OpenAI SDKs. There is no local process to run; the server is remote and authenticated with your key.

Other MCP servers · Engineering and operations

• GitHub MCP server
• AWS CloudWatch MCP server
• Sentry MCP server
• ServiceNow MCP server
• Okta MCP server
• Twilio MCP server