← Back to home
Comparison · DevOps

Sanity vs GitHub

Side-by-side trajectory, velocity, and editorial themes.

S
Sanity
DEVOPS
5.0

Sanity is quietly wiring its CMS to be operated by agents as much as by humans.

◆ Current state

Sanity is shipping on several fronts in parallel: a maturing MCP server and agent tooling, a Media Library growing real asset-management depth, and steady Studio and SDK ergonomics. The recent run is incremental but coherent — richer Media Library metadata and reference tracking, searchable reference fields, and a stream of MCP tool fixes. Nothing here reshapes the product; it is compounding polish on an already broad platform.

◆ Where it's heading

The clearest theme is agent-operability. The MCP server, a skills-install CLI command, agent-focused doc quickstarts, and copy-paste commands 'for humans and agents' all point at Sanity treating AI coding agents as a first-class way to drive the CMS. In parallel, Media Library is being built out toward a full DAM, and @sanity/presets is trimming schema boilerplate.

◆ Prediction

Expect the MCP and agent surface to keep expanding and Media Library to keep gaining DAM-grade features; the presets package suggests more ready-made schema building blocks ahead.

GitHub logo
GitHub
DEVOPSCOLLAB
10.0

Every new Copilot capability now ships with an enterprise dial bolted to it.

◆ Current state

GitHub is shipping on three fronts at once: Copilot model and UX, code-security scanning, and enterprise governance. The past two weeks lean hard toward giving org admins granular control, from per-repository code-quality targeting and mandated OpenTelemetry export to per-user budget visibility, while Copilot keeps absorbing frontier models. Security tooling is maturing from raw detection breadth toward operational clarity and internal-only workflows.

◆ Where it's heading

The platform is converging on a governed AI-development surface where each Copilot feature arrives paired with an admin control and a telemetry hook. Security scanning is bending toward AI-era threats like prompt injection and toward enterprise-internal patterns such as innersource advisories. The admin-control and observability surface is expanding in lockstep with every model addition.

◆ Prediction

Expect the next moves to continue the pattern visible here: more governance around Copilot (budgets, policy, telemetry) landing alongside the next frontier-model onboarding, rather than a standalone new product.

See more alternatives to Sanity
See more alternatives to GitHub