Recently shipped
Coderbuds Standards and the pre-flight fit check
- Coding agents call assess-change-fit before opening a pull request, and get a verdict against how your team actually ships
- Two standards judged before review starts: pull request size, and scope coherence — the distinct outcomes a reviewer has to approve
- The size standard calibrates on your own merged work, and proposes moving itself when the overrides say it is wrong
- Every check is recorded and linked to the pull request it became, so followed and ignored advice both get receipts
- Standards scoreboard on the Agents page, and a standards line in the weekly summary
The org map for coding agents
- get-team-context gives an agent the live picture of how your repositories fit together, scoped to the one it is working in
- Dependency edges derived from package.json and composer.json manifests, including monorepo subdirectories
- Co-change coupling ranks the repositories that keep shipping together
- Drift check flags rules files still pointing at archived repositories
- link-repositories teaches the map the relationships it cannot observe on its own
MCP: one-click connect, and tools worth calling
- Connect over OAuth 2.1 in one click, with headless tokens kept as the fallback for CI
- Twenty-nine tools covering standards, the org map, pull requests, deploys, DORA and insights
- get-my-status: your open pull requests, the reviews you could unblock, and your merged-but-unshipped work in one call
- get-repo-readiness answers whether a repository is set up for agents at all
- Two-minute copy-paste setup for Claude Code, Cursor and Codex, and adoption telemetry so you can see who is actually using it
Deploy tracking, and who actually deployed
- Deploys attributed to the person who triggered them rather than whoever owns the token
- An inbound deployment webhook for platforms that never emit a provider event
- Guided deploy-hook setup that says plainly when a hook is recording nothing
- "Who deployed to production this week, and how much" — with failed deploys surfaced per person
- Deploys can be triggered from Slack or from an agent over MCP
The delivery loop, in Slack
- A ready-to-deploy queue that reposts when it gets buried, with a Deploy now button
- /review works on any pull request, including the ones your agents wrote
- One weekly pulse post in place of five separate agent summaries
- Ask Coderbuds in the channel, right after the deploy
- Request a review from Slack, from the dashboard, or from an agent
The dashboard rebuilt around the loop
- Nine tabs cut to four, with the Overview reduced to Today, Delivery and Health
- A setup checklist derived from what we can actually see, not from a list you tick yourself
- Pull request flow, a six-month shipped trend, and a profile for each engineer
- Empty sections say why they are empty instead of reporting health from no data
AI adoption, measured honestly
- Agents declare their own contribution over MCP with report-ai-usage
- Those declarations grade the detector rather than being quietly blended into it
- Reported as team-level trends only — never as a verdict on who wrote what
Reports and the partner API
- Board, leadership and engineering templates with eleven selectable sections
- Custom date ranges and repository scoping, exported as PDF, Markdown or straight into Slack
- A REST endpoint at /api/v1/metrics/monthly for building your own dashboard
- Ticket delivery mix and commitment reliability read from Linear
Agents, insights and integrations
- The coaching agent fleet, with outcome tracking and a live activity feed
- Fifteen insight analyzers across the SPACE dimensions, with personal and team insights
- GitHub, BitBucket, Jira, Linear, Slack and Microsoft Teams connections
- DORA metrics with automatic deployment-to-pull-request association
- Weekly pulse surveys with Slack reminders and per-member response tracking
Coming next
The fast lane
In progressA change that fits gets routed to the reviewer best placed to approve it and merged fast. Standards are only worth keeping if keeping them is quicker than not.
Receipts you can check
In progressPull requests that went through the fit check compared against the ones that did not — size, merge time, and how often the advice was taken. Published, not asserted.
Teaching the org map
PlannedFacts about your system an agent cannot observe, recorded with who said them and when they stop being true. Sparingly, so the map never becomes another wiki to maintain.
Weekly changelog email
PlannedThe weekly summary your Slack channel already gets, delivered to the inboxes of the people who do not live in Slack.
Earned autonomy
ExploringAgents acting without asking, but only on the specific actions a team has already been approving by hand. Autonomy follows evidence rather than leading it.
What we're not building
A roadmap is only useful if it rules things out. These come up often enough to be worth answering here.
A settings page for every threshold
The size standard calibrates on your team's own merged work and proposes moving when your overrides disagree with it. You get one switch, not a settings farm to tune.
AI code review as the headline
Reviewing the diff is a crowded, well-served problem, and it is not where work gets stuck. We keep review as a feature and spend our effort on landing the change.
An autonomy dial per agent
One team-level posture governs the whole fleet, so there is a single answer to what agents are allowed to do here. Per-agent switches turn that into a matrix nobody can hold in their head.
Have a feature request?
We'd love to hear what you'd like to see next. Reach out and help shape the roadmap.