Blog
Practical insights on agentic software delivery
Featured
How to build an AI software factory: a roadmap in four stages
A four-stage roadmap from a first readiness check to a factory that runs every day. See what each stage needs, when it starts and what it delivers.
SWE-in-a-team: a coding benchmark for software factories
Our SWE-in-a-team benchmark graded thirteen coding agents and models in an SDLC loop. An Open-Weight builder resolved every ticket at half the price.
Agentic SDLC, defined: a gated loop from issue to verified pull request
An agentic SDLC takes each change through plan, build, CI, review, preview and test, and sends rejected work back. In 260 missions, the loop fixed 40 of 46.
Latest
Codex vs Claude Code in 2026: 120 missions on the same 20 tickets
On 20 tickets in one codebase, codex resolved 55 of 60 for $3.63 per resolved ticket and claude-code 58 of 60 for $4.69. Each ran three models.
AI prototype to production: get a reviewed, tested pull request
Take an AI prototype to production through a plan, your CI, a review, a preview and a test. A workspace for non-engineers is in early access.
Claude Code in Slack: start it from a thread, and join early access to SHIP
Claude Code works in Slack today through the Claude app. SHIP's loop that turns a Slack thread into a tested pull request is in early access.
Claude Code on Jira: connect it today, and assign tickets to SHIP
Connect Claude Code to Jira today with Atlassian's MCP server. Then assign a ticket to SHIP and get a pull request that passed your CI, a review and a test.
Codex on Linear: assign an issue, get a tested pull request
Assign a Linear issue to SHIP. Codex builds it in a sandbox, and your CI, the Reviewer and QA check it before the pull request reaches you.
Run Claude Code headless on real tickets, in a sandbox
Run Claude Code without a terminal in an isolated sandbox that never holds your key, with CI, review and QA around it. 77 of 80 benchmark missions resolved.
Run Codex headless with codex exec, on real tickets
Run Codex in an isolated sandbox that never holds your OpenAI key, with CI, review and QA around the run. Codex builders resolved 55 of 60 benchmark missions.
What is shadow AI, and what it looks like in software development
Shadow AI is AI use without IT or security approval. In software, it is code built with AI outside your review and deploy pipeline.
Vibe coding security: the checks between a prototype and production
Vibe coding security comes from checks the author did not write. See the risks the security guides name, and the checks SHIP runs on every change today.
What is a dark software factory, and where do people still stand?
A dark software factory ships code that no person reads, and only machines check it. See where the term comes from and what 260 agent missions show.
What is agentic coding, and what makes its output safe to merge
Agentic coding means AI agents plan, edit, test and fix code for a task. In 260 SHIP missions, gates outside the agent sent 46 runs back to be fixed.
What is governed vibe coding, and who decides what merges
Governed vibe coding lets people outside engineering build software with AI, while engineering controls where it runs, who reviews it and what merges.
Connect GitHub's and Linear's MCP servers to your coding agents
Bind GitHub's MCP server with a scoped token and Linear's with an OAuth sign-in. Keep review and QA read-only, and see each call on the run.
MCP servers for coding agents: bind them once, keep the keys out of the sandbox
Give your planner, builder, reviewer and QA agents the MCP servers your repository already declares, with credentials that never enter the agent's sandbox.
Run the pi coding agent headless on real tickets, in a sandbox
Run the pi coding agent as SHIP's builder in an isolated sandbox, on an open-weight model through your own OpenRouter key, with review and QA around it.
Codex code review for Claude Code: set up a second-provider reviewer
Keep Claude Code on the builder and put Codex on the reviewer, in ship.yml or for one mission, so a second provider reads every diff before you do.
Claude Code on Linear: assign an issue, get a tested pull request
Assign a Linear issue to SHIP and Claude Code plans it, builds it, clears your CI, reviews it and tests it on a preview before the pull request reaches you.
What is an AI software factory, and how to build one with Claude Code
An AI software factory takes each issue through agent stages and verification gates to a pull request a person accepts, and records what every change cost.
Introducing SHIP - the agentic engineering platform you need
SHIP is here to help engineering teams deliver more with AI, control costs, and keep the quality practices that make software reliable.