Item detail

mindfold-ai/Trellis

Trellis is an AGPL-3.0 engineering framework for AI coding that persists specs, tasks, and project memory into your repo so any coding agent works to your engineering standards instead of starting from scratch every session. 10.5k stars, npm-published as @mindfoldhq/trellis, multi-platform (Claude Code, Codex, Cursor, OpenCode), and a documented real-world-scenarios doc set covering bug-fix, featu

Score8.3
Popularity88.0
Riskconditional
TierGold
Score breakdown
Usefulness8.0
Novelty8.0
Momentum9.0
Maturity8.3
Open-source/build8.4
Evidence7.2
Workflow potential9.8
Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for engineering teams that want every AI coding agent — Claude Code, Cursor, Codex, OpenCode — to operate against the same specs, tasks, and conventions instead of each agent rediscovering them. Install via npm, point Trellis at a repo, and let it record the spec/task/memory so successive runs inherit context.

Who should use it

engineering teams that want every coding agent — Claude Code, Cursor, Codex, OpenCode — to operate against the same specs and conventionstech leads who want shared specs and task lists in-repo instead of buried in Notion or Slack threadsopen-source maintainers who want AI agents to follow documented contributor guidelines without re-explaining them every sessionteams standardizing on multiple AI coding tools who need one consistent harness across themanyone tired of re-prompting the same project context to a fresh agent session

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

AGPL-3.0 license is a real consideration for closed-source commercial products — internal and open-source use is fine; persists specs/tasks/memory inside the repo, so contributors will see (and can edit) the harness state.

Evidence links

Closest alternatives / related signals

agent-frameworkscoding-agentsengineeringspecstask-memoryclaude-codecursorcodex