Open sourceTypeScript · ≈86 stars · Apache-2.0
Quest
An open-source, self-hosted, safety-focused AI workbench.
Quest is Electric Capital's open-source, self-hosted, safety-focused AI workbench — a harness built around permissioning and trust for agents doing real work. Early-stage (Sep 2026) but backed by a firm that funds much of open-source infrastructure.
Activity
- Last push
- · 2d ago
- Latest release
- v0.5.0
· 4d ago - Last 90 days
- 7 releases
- · 4d ago
Highlights Doc notes: a Quest Doc can now have its own instructions and memories. The agent reads them before it works on the doc, and can suggest new ones for you to approve. Dreaming: Quest can look over your saved memories in the background and suggest cleanups. Nothing…
- · 3w ago
Highlights Thinking level picker: in the model menu, the thinking level list no longer vanishes when you move your mouse over to it. You can actually pick a level now. Change previews: when the agent edits a doc, skill or routine, the preview now highlights just the words that…
- · 3w ago
Highlights Quest Tables: you can now keep structured tables in Quest. Make them, filter them, edit the columns, and look back through their history. The agent can read and change them too, and asks you first before it writes. Memory suggestions: Quest now suggests memories on…
- · 5w ago
Highlights Quest Docs: in-app markdown docs with sharing, editing, history, per-doc approval for agent writes, and per-project access to public projects' docs Conversation workspaces: chat:// and proj:// file spaces, Chat Files / Project Files cards, Copy/Move between spaces…
- · 7w ago
What's Changed README: link the Quest install video in Quick Start by @puntium in https://github.com/electric-capital/quest/pull/19 Bake headless LibreOffice + fonts into the sandbox images as the document converter by @cody-the-code-bot in…
Checked against GitHub on Nov 21, 2026. Release notes are shortened to plain text.
Neighbours on the workbench
JavaScript · ≈308k stars · MIT
affaan-m/ECC
Agent harness performance optimization system: skills, instincts, memory and security for coding agents.
Maintained by Affaan Mustafa
JavaScript · ≈76k stars · MIT
cloudflare/security-audit-skill
Cloudflare's skill for multi-phase security audits with independently verified findings.
Maintained by cloudflare
TypeScript · ≈28k stars · MIT
promptfoo/promptfoo
Test your prompts, agents and RAG — plus red-teaming.
Maintained by promptfoo
code-scan-action-0.2.3releasedConnections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
Maintained by · 1
- Electric CapitalOrganisation
