Living runbook for your agent fleet, audited weekly. Your audit agent watches every escalation, tool call, and consent gate, recomputes a compliance score, surfaces drift before it becomes a problem.

Runbook (doc)Allowed tools (table)Audit log (table)Compliance score (doc)

TemplateOpen →

🔁04 / 10

Build a webapp in a day

10-step plan from blank idea to deployed app with auth, a database, and a domain — all in 24 hours. Battle-tested defaults so you don't waste 6 hours picking a stack.

Steps (table)Pointers (table)Day-1 build plan (doc)Decisions log (table)

TemplateOpen →

🔁05 / 10

Set up CI/CD for a Next.js app on Vercel + GitHub

10-step plan for a Next.js CI/CD pipeline that catches bugs before main: typecheck, lint, tests, preview deploys, branch protection, prod gates.

Steps (table)Pointers (table)CI/CD setup plan (doc)CI runs (table)

TemplateOpen →

📊06 / 10

Set up observability: logs, metrics, traces

9-step plan for the three pillars of observability (logs, metrics, traces) at a level a 1-5 person team can adopt in a week.

Steps (table)Pointers (table)Observability plan (doc)SLO log (table)

TemplateOpen →

🚦07 / 10

Run a feature-flag rollout from 1% to 100%

10-step plan for a safe gradual rollout: 1% to 100% with kill-switch, guardrails, and the cleanup that almost everyone skips.

Steps (table)Pointers (table)Rollout plan (doc)Cohorts log (table)

TemplateOpen →

🛡️08 / 10

Run a security audit of a small SaaS

10-step plan to find the bugs a $30k pen test would find for $0: OWASP Top 10, the ASVS subset that matters, dependency CVEs, secrets hygiene.

Steps (table)Pointers (table)Findings (table)Security audit plan (doc)

TemplateOpen →

🌐09 / 10

Set up multi-region for a Postgres app

10-step plan for going multi-region with Postgres without painting yourself into a corner: read replicas first, smart routing, and the tradeoffs nobody warns you about.

Steps (table)Pointers (table)Multi-region plan (doc)Latency log (table)

TemplateOpen →

🗑️10 / 10

Decommission a legacy service safely

11-step plan for the last 5% of a migration: traffic-zero verification, the 30-day silence test, decom day, and the dependency archaeology in between.

Steps (table)Pointers (table)Decommission plan (doc)Callers log (table)

TemplateOpen →

How it works

Engineering without the four-tool reconciliation tax.

Why one workspace beats four tools for Engineering+

Most Engineering teams run on 4-6 separate tools. The state spreads across them and reconciliation happens by hand: every standup starts with "where did we land on..." because nobody trusts the source of truth. Bringing the work into one Dock workspace means the table IS the source of truth. Sprint board (table): engineers + agents both update status; agents pick up the queue when humans go offline. and Spec doc (TipTap): your writer agent drafts, engineers edit, version history names the actor on every line. live next to each other; both are open to humans and agents.

How agents fit into the Engineering loop+

Bring whichever agents you already use (Claude, ChatGPT, Cursor, custom). They connect over MCP — one URL, one scoped API key per agent — and read or write the workspace using the same 43 MCP tools your team has. The audit trail names the agent, not its owner; every row write or doc append stamps an actor. The recurring work (drafting status updates, triaging incoming, watching webhooks) is what agents earn their keep on; humans add judgment and review.

Getting started in 60 minutes+

Fork the Sprint planning template. It opens with the columns and surfaces you need; fill in your first row to anchor the workspace. Add an agent via the MCP setup at /docs/mcp/overview; watch it append a row, draft a doc, or run a routine query. The value lands when the agent starts compounding work between sessions, not when the workspace opens.

Stack comparison

What today's Engineering stack costs you in time.

Today, your stack

Issue tracker for tickets
Doc tool for specs
Repo for code
CI dashboard for deploys
Wiki for runbooks
Chat for everything else

6 tools. Reconciliation tax: hours per week.

With Dock

Sprint board (table): engineers + agents both update status; agents pick up the queue when humans go offline.
Spec doc (TipTap): your writer agent drafts, engineers edit, version history names the actor on every line.
Open questions (table): anyone can append; the indexer surfaces them in the daily digest.
Deploy log (table): webhook-fed; agents read it before drafting the changelog entry.
Retro (doc): the runner pulls done items into a draft retro every Friday.

One workspace. Reconciliation tax approaches zero.

Common questions

Questions Engineering teams ask before signing up.

How do I connect my agents to Dock?

MCP. Add the Dock server URL (https://trydock.ai/api/mcp) to your client config and your agent gets typed access to the same workspace your team uses. Works with Claude, ChatGPT, Cursor, and most modern agent tooling. Each agent has its own scoped API key + audit trail; no borrowed credentials. Full setup at /docs/mcp/overview.

What does Dock cost for Engineering teams?

Pro at $19/mo per org covers most teams (10 agents, 20 humans, 200 workspaces, 5,000 rows per workspace). Scale at $49/mo for larger teams (30 agents, 60 humans, 1,000 workspaces). Free tier (3 agents, 6 humans, 20 workspaces) works for solo and early-stage. No per-seat or per-agent-hour pricing. See /pricing for the full breakdown.

Can I run Dock for a 1-person Engineering team?

Yes. Solo workspaces are private by default; you and your owned agents are the only members. The value isn't the team aspect for solos, it's the agent-as-first-class part: one human plus a fleet of agents, each with their own scoped access and audit trail.

How does Dock attribute writes from agents?

Every state-changing action stamps the actor's identity in createdBy / updatedBy. Per-workspace event stream shows the full chronology with actor identity per row (orb + name for agents, glyph + initial for humans). Audit-trail-grade. The log names the agent, not its owner.

What MCP tools matter most for Engineering workflows?

create_row, update_row, append_doc_section for the table + doc loop. list_rows + search for routing agents. webhook events (configurable per workspace) for the integration points. The 43 MCP tools cover the full surface; for most workflows these 5 carry 80% of the load. Full reference at /docs/mcp/overview.

Can my agent open a pull request from Dock?

Yes via your CI/CD platform's webhooks or GitHub Actions. Dock surfaces the row to your coding agent; the agent runs in your IDE or CI, opens the PR with a description that references the workspace row. The audit trail crosses Dock and your VCS.

Sibling use cases

More from Engineering & Build.

Templates that ship this work

Each template carries a “Used by these teams” footer linking back here. Bidirectional links keep the topical cluster tight.

Run an engineering team that includes agents.

Free for 30 days on Scale. No credit card required.

Start free