Scroll

Production SDLC,
while you sleep.

0
Manual handoffs
5
Harnesses supported
Tickets processable

Trellis is a bulk execution framework for shipping large-scale software on a schedule. Install it once, wire it to cron, and your backlog ships itself.

It runs as a skill inside your agent harness. One install, six commands. Each command is stateless and idempotent. Running one twice produces no duplicate work, making the pipeline safe to run on a schedule.

Research before implementation. Tests before review. Comments resolved before merge. The full SDLC runs unattended.

One copy.
Zero build steps.

1

Clone or copy

Clone the repository, or copy the skills/trellis folder directly into your agent's skills directory.

2

Set environment variables

Add LINEAR_API_KEY and GH_TOKEN to your environment. Each command lists which it requires.

3

Run your first command

Invoke /trellis vision-roadmap to generate your first roadmap and backlog, or drop any single command into an existing workflow.

Claude Code
$ git clone https://github.com/drewdoebereiner/trellis.git
$ cp -r trellis/skills/trellis ~/.claude/skills/
Commands
/trellis vision-roadmap
Read the repo, ask 7 clarifying questions, produce a strategic roadmap, create Linear backlog tickets
/trellis research-backlog
Fetch all backlog tickets, dispatch parallel research agents, post structured findings as comments
/trellis dev-backlog
Pull unstarted tickets by priority, implement each on its own branch, open PRs, move to In Review
/trellis write-unit-tests
Discover the test framework, find coverage gaps from git diff, write pattern-matched tests
/trellis bulk-pr-review
Map PR dependency layers, dispatch parallel review agents, post verdicts directly to GitHub
/trellis fix-pr-comments
Find PRs with open feedback, fix every issue, push fixes, reply to each thread with the commit SHA
A trellis doesn't control the plant. It gives it structure to grow along.

Structure

The framework defines skills and shape. Agents traverse a lattice of composable capabilities, finding the path of least resistance through your codebase.

Autonomy

Six sub-commands cover the full SDLC loop, from roadmap to merged PR, with no manual handoffs between stages. Wire it to cron once and step away.

Scale

Most agent tooling handles one feature at a time. Trellis processes your entire backlog, PR queue, and test coverage gaps in a single automated pass, every night.

Built for production scale. Not proof of concept.

01

Bulk, not one-shot

Most agent tooling helps you write one feature at a time. Trellis processes your entire backlog, PR queue, and test coverage gaps in a single automated pass, every night if you want.

02

Direct API, not MCP

Trellis calls Linear and GitHub directly via curl rather than through MCP servers. In bulk runs with 5 to 25 parallel subagents, MCP overhead compounds. Direct calls return exactly what is needed, leaving more room for actual work.

03

Harness-agnostic

Skills are plain markdown files with no build step and no dependencies. Any agent harness that can load and execute skills can run Trellis. Claude Code, Cursor, Gemini CLI, and Codex all work out of the box.

04

Composable by design

Run the full SDLC loop end-to-end, or drop a single sub-command in wherever you need it. Each stage is independently useful and can run without the others. The pipeline is additive, not monolithic.

Six commands.
The complete loop.

Each command maps to a stage of the software development lifecycle. Run them in sequence for a full automated sprint, or drop one in wherever you need it. All invoked as /trellis <command>.

01
/trellis vision-roadmap
Roadmap and Backlog Generation
Reads the repo, asks 7 clarifying questions, produces a strategic roadmap document, and creates Linear backlog tickets via GraphQL. The starting point for any new sprint.
LINEAR_API_KEY
02
/trellis research-backlog
Parallel Ticket Research
Fetches all backlog tickets, dispatches parallel research agents, and posts structured findings as comments. Skips tickets it has already commented on, so it is safe to run daily.
LINEAR_API_KEY
03
/trellis dev-backlog
Bulk Implementation
Pulls unstarted tickets by priority (default 10, configurable), implements each on its own branch, opens PRs targeting dev, and moves tickets to In Review automatically.
LINEAR_API_KEY GH_TOKEN
04
/trellis write-unit-tests
Coverage from Diff
Discovers the test framework from the codebase, identifies coverage gaps from git diff, and writes pattern-matched tests for every changed file. No configuration required.
no env vars
05
/trellis bulk-pr-review
Parallel PR Review
Maps PR dependency layers, dispatches parallel review subagents per layer, and posts APPROVE / REQUEST CHANGES / BLOCKED-CI verdicts directly to GitHub. Every open PR reviewed in one pass.
GH_TOKEN
06
/trellis fix-pr-comments
Automated Comment Resolution
Finds all open PRs with CHANGES_REQUESTED or unresolved threads, fixes every issue, pushes to the branch, and replies to each thread with the resolving commit SHA.
GH_TOKEN

Set it up once.
The loop runs itself.

Because every command is stateless and idempotent, the entire pipeline is safe to run on a cron schedule. No state files. No locks. No cleanup. Running a command twice never creates duplicate work.

Trellis tracks its own progress through Linear ticket states and GitHub PR metadata. Research agents skip tickets they have already commented on. dev-backlog only picks up unstarted tickets. fix-pr-comments only touches PRs with open feedback.

Wire the pipeline into cron once. Implementation runs every weeknight, PR review every afternoon, comment fixes before standup. No manual triggers needed.

crontab
0 8 * * 1-5 /trellis research-backlog
0 21 * * 1-5 /trellis dev-backlog
0 14 * * 1-5 /trellis bulk-pr-review
0 9 * * 1-5 /trellis fix-pr-comments

Works where your agents already live.

Trellis is harness-agnostic. Copy the skills folder into your agent's skills directory and the commands are available immediately. No build step, no configuration beyond environment variables.

Claude Code ~/.claude/skills/
Cursor .cursor/skills/
Gemini CLI ~/.gemini/skills/
OpenAI Codex CLI ~/.codex/skills/
Any harness with skill support check your docs