What you can do with Codefresh

Built for

Platform Engineers, DevOps Teams, Release Managers, SREs, Kubernetes Operators

Example workflows

Recent build failures

Release managers triage failed builds before standup.

Try this

In Codefresh, show the 10 most recent workflow builds and summarize which pipelines failed in the last 24 hours.

Pipeline inventory

Platform teams audit stale CI configuration regularly.

Try this

List all Codefresh pipelines grouped by project and flag any that have not run in the past week.

Context health check

On-call engineers verify wiring before rerunning deploys.

Try this

Read my Codefresh integration contexts and execution contexts — note any missing Kubernetes or registry connections.

Audit trail

Compliance reviews need actor and timestamp detail.

Try this

Pull the latest Codefresh audit events and summarize who changed pipeline or context settings this week.

Context to know first

What is Codefresh MCP?

Codefresh MCP connects your Codefresh account to AI assistants so you can inspect pipelines, builds, contexts, agents, and analytics from chat without copying dashboard exports.

Can agents trigger builds or edit pipelines?

This integration is read-only. Agents can inspect CI/CD state and analytics but cannot trigger builds, create pipelines, or mutate account settings.

Why do build reads return download links?

Build history responses include full pipeline YAML per run. Large pages spill to a presigned download link instead of inline JSON — use a small limit (10–25) for interactive triage.

AI Skill
SKILL.md

Domain knowledge for Codefresh — workflow patterns, data models, and gotchas for your AI agent.

Codefresh

Codefresh automates build, test, and deploy workflows for cloud-native applications with GitOps and Kubernetes-native delivery.

Operational order

  1. Account settings — obtain account_id for team-member lookups.
  2. Projects and pipelines — navigate project-scoped pipeline names before build history.
  3. Builds — inspect workflow runs; paginate sequentially with a small page size.
  4. Contexts and execution contexts — verify Git, registry, and runtime cluster wiring.
  5. Analytics — read analytics metadata before requesting a named report.

Data Model

  • Account settings — security and general preferences; provides account_id.
  • Account users — team members for a given account_id.
  • Agents — runtime agents connected to the platform.
  • Analytics metadata / reports — report catalog and per-report payloads.
  • Audit — compliance and change events.
  • Builds — workflow run history (YAML-heavy payloads).
  • Contexts — Git, registry, and cloud integration contexts.
  • Execution contexts — runtime clusters and hybrid environments.
  • Helm repos — chart repositories for Kubernetes deployments.
  • Pipelines — CI/CD pipeline definitions and status.
  • Projects — project groupings for pipelines.
  • Step types — available pipeline step templates.

Gotchas

  • account_id required for team-member reads — resolve from account settings first.
  • Build pagination: sequential integer pages in one session — omit page, then page=2, page=3 when pagination.nextPage is true. Skipping ahead fails with PAGINATION_BAD_REQUEST_ERROR.
  • Build payload size: each build may include full pipeline YAML; keep limit at 10–25 on build history reads.
  • Read-only integration: this bundle covers inspection and analytics only — pipeline triggers and CRUD are not exposed here.

Tools in this Server (13)

Codefresh Get Account Settings

Read Codefresh account settings including security preferences and general configuration. Use this first to obtain account_id for team-member lookups.

Codefresh Get Account Users

Read team members and access for a Codefresh account. Requires account_id from account settings. Returns user records for the specified account.

Codefresh Get Agents

Read runtime agents connected to your Codefresh account. Use to inspect agent status, versions, and metadata across Kubernetes or hybrid runners.

Codefresh Get Analytics Metadata

Read analytics metadata describing available Codefresh reports, dimensions, and date-range options. Use before pulling a named analytics report.

Codefresh Get Analytics Reports

Read a specific Codefresh analytics report by report_name. Discover valid report_name values from analytics metadata first.

Codefresh Get Audit

Read Codefresh audit events for compliance and change tracking. Returns event types, timestamps, and actors. Paginate with page and pageSize.

Codefresh Get Builds

Read Codefresh workflow builds. Paginate sequentially: omit page on the first call, then page=2 when pagination.nextPage is true on the prior response...

Codefresh Get Contexts

Read integration contexts in Codefresh (Git, Docker registries, cloud providers, Kubernetes clusters). Use when wiring pipelines or debugging authenti...

Codefresh Get Execution Contexts

Read execution contexts that define where pipeline steps run (runtime clusters, hybrid environments). Use to distinguish deployment targets.

Codefresh Get Helm Repos

Read Helm chart repositories configured in Codefresh. Returns repository names, URLs, and authentication settings for Kubernetes deployments.

Codefresh Get Pipelines

Read CI/CD pipelines in Codefresh. Returns pipeline identifiers, names, and status metadata. Paginate with offset and limit when the catalog is large.

Codefresh Get Projects

Read Codefresh projects that group pipelines and repositories. Use to navigate project-scoped pipeline names and workspace organization.

Codefresh Get Step Types

Read available pipeline step types and their configuration options. Use when designing or modifying pipeline YAML.

Frequently Asked Questions

What is Codefresh MCP?

Codefresh MCP connects your Codefresh account to AI assistants so you can inspect pipelines, builds, contexts, agents, and analytics from chat without copying dashboard exports.

Can agents trigger builds or edit pipelines?

This integration is read-only. Agents can inspect CI/CD state and analytics but cannot trigger builds, create pipelines, or mutate account settings.

Why do build reads return download links?

Build history responses include full pipeline YAML per run. Large pages spill to a presigned download link instead of inline JSON — use a small limit (10–25) for interactive triage.

How does build pagination work?

Build pages are sequential within one session. Start without a page number, then request page 2 only when the prior response shows pagination.nextPage=true. Skipping pages returns an error.

What credentials does Codefresh MCP need?

Connect with a Codefresh API key or OAuth from your Codefresh account settings. The key must have read access to the resources you want agents to inspect.

How do I connect Codefresh to my AI agent?

Add the MCPBundles server URL to your MCP client configuration (Claude Desktop, Cursor, VS Code, etc.). The URL format is: https://mcp.mcpbundles.com/bundle/codefresh. Authentication is handled automatically.

How many tools does Codefresh provide?

Codefresh provides 13 tools that can be called by AI agents, along with a SKILL.md that gives your AI agent domain knowledge about when and how to use them.

What authentication does Codefresh require?

Codefresh uses API Key or One-click sign in. Codefresh requires credentials. Connect via MCPBundles and authentication is handled automatically.

Setup Instructions

Connect Codefresh to any MCP client in minutes

MCP URL
https://mcp.mcpbundles.com/bundle/codefresh

One-click install:

The link prefills the Add custom connector dialog — you still review the values and click Add, then Connect to complete OAuth.

Or add manually

  1. Open claude.ai → Settings → Connectors.
  2. Click the + button and choose Add custom connector.
  3. Set Name to Codefresh and paste the MCP URL into Remote MCP server URL.
  4. Click Add. Codefresh will appear under Not connected — select it and click Connect to complete OAuth.
Name: Codefresh
Remote MCP server URL: https://mcp.mcpbundles.com/bundle/codefresh
Authentication: OAuth

Custom connectors at claude.ai require a paid Claude plan (Pro, Max, Team, or Enterprise).

Ready to use Codefresh?

Sign in to connect your credentials and start running tools from the chat.