← CHASSIS USE CASE · ENGINEERING TEAMS

Chassis for engineering teams.

The PM Drive for engineering teams that don't want to pay for Linear plus a Slack-AI plus a separate retro tool. GitHub bidirectional sync is in the box. Forge AI ships at every tier. Cycles, burndown, points, priority — the table-stakes work today.

Bundled with the TECH exec hire ($149/mo) — a CTO-shaped exec that operates Chassis, Atlas (knowledge management), and Beacon (technical SEO + page audits) as one engineering surface.

The job.

Engineering teams ship features against tickets, run sprints, review PRs, handle on-call, gate deploys. The work is bursty and parallel: a sprint of 30 tickets while two infra migrations and a security review run alongside. The PM tool has to track all of it, including the work that doesn't have a clean GitHub issue yet.

Most teams duct-tape three tools together: a PM tool (Linear / Jira), a knowledge base (Confluence / Notion), and a chat-first AI thing (Slack AI / random Copilots) to summarize what shipped. The seams are where the leakage lives. Sprint retro happens in a Google Doc. The PR list never makes it back to the PM tool. The tickets don't know who actually owns the on-call rotation. Half the AI summaries are about the wrong sprint.

Chassis is the project layer that ties those pieces together inside the cockpit. Tasks know about their PRs. Cycles know what shipped. Forge writes the retro from the activity log, not from your memory at 5pm Friday. The same workspace also tracks the non-code work: vendor migrations, security reviews, SOC 2 evidence collection, the architecture spike that hasn't grown a repo yet.

It is not a Linear replacement for a 200-person org with five years of muscle memory. It is the PM Drive for the team of 3–20 engineers that wants one tool, GitHub-native, with the AI workflows already wired in.

How Chassis fits the engineering workflow.

Link a github_repo. Engineering view turns on.

Set the field on a project. Sync banner appears. Tasks gain GitHub issue + PR URL fields. Create-GitHub-issue button shows up on tasks not yet linked. Auto-close-on-done becomes available. One toggle moves a project from operations-shaped to engineering-shaped.

Bidirectional sync — import, reconcile, export, close.

Import pulls open issues + open PRs into Chassis tasks. Reconcile updates task status when GitHub state changes (PR merged → done, issue closed → done). Export creates a fresh GitHub issue from a Chassis task and persists the URL back. Close auto-closes the linked GitHub issue when a Chassis task flips to done. Auth via the Merkava GitHub App or a personal access token.

Forge subtask suggestions, on every task.

Hit Suggest. Forge proposes 3–5 concrete subtasks with descriptions, drawing on the task body, the project, and your venture profile. Auto-creates them as children. Delete the ones you don't want. Most useful on the "implement X" tickets that are too big to estimate cleanly.

Cycles + burndown + the first-sprint wizard.

Cycles with start + end dates, points (Fibonacci), priority. Burndown chart per cycle. The "Run first sprint" wizard reads your starter project, sets the dates, picks a backlog, seeds the kanban — the empty state that usually takes a half-day in a new tool ships in under a minute.

Auto-written retros at end-of-cycle.

Close a cycle. Forge writes the retro: what shipped, what slipped, keep / cut / start. Pulls from the activity log so it's grounded in what actually happened. Edit the draft, paste into Slack or your weekly review doc, move on.

Owners are real people, not just emails.

Task owners are Crew employee records — the same record HR, payroll, performance, and onboarding all reference. When the on-call rotation flips, every Chassis task assigned to "the on-call person" updates without a re-key. When an engineer leaves, ownership reassigns at the Crew layer and Chassis follows.

Workflow walkthrough — a typical sprint.

Concrete five-step example. Sprint 14 of an early-stage product.

  1. Open the sprint. Hit Run first sprint (or New cycle on subsequent ones). Forge sets dates two weeks out, pulls a backlog from the linked github_repo project (open issues, weighted by priority), assigns Fibonacci estimates from the prior sprint's velocity. Review, adjust, confirm.
  2. Import open issues + PRs. Sync banner. Hit Import. Open GitHub issues become Chassis tasks (linked, with original URL). Open PRs attach to existing tasks where the title matches; new PRs get fresh tasks. Idempotent — re-runs don't dupe.
  3. Daily kanban + standup. Engineers move their tasks across To-do / Doing / Blocked / Done. Activity log captures every move + comment. Reconcile runs daily (or on-demand) to flip done-state when PRs merge.
  4. Block-detection. Tasks that have been in Blocked > 7 days surface to the engineering lead's inbox. Drop, escalate, unblock — three buttons, audit-logged.
  5. Close cycle, get auto-retro. Hit Close. Forge writes the retro grounded in the activity log: what shipped (PRs merged + tasks done), what slipped (carried-over tasks), keep / cut / start. Edit, ship to Slack, move on. Burndown + velocity update for the next cycle's wizard.

The other Drives that come with this.

Engineering teams rarely run on a PM tool alone. Three Drives in the TECH bundle that pair with Chassis.

PRICING

Pricing for engineering teams.

  • Standalone Chassis — $49/mo. Just the PM Drive, no exec. Right if you already have your own engineering lead and only want the project layer.
  • TECH exec — $149/mo. CTO + 3 specialists. Includes Chassis + Atlas + Beacon. Right pick for most engineering-led teams.
  • Full C-Suite — $599/mo. All five execs, every Drive in the box.
  • 7-day free trial on first install. Cockpit itself is free. Annual billing saves 20%. Per-tenant pricing — no per-seat surprise as the team grows.

Run engineering on Chassis.

Sign up for Merkava, install Chassis, link a GitHub repo, hit Import. Sprinting in 5 minutes.

Get started Back to Chassis See pricing