Latest release v2026.1012.0 is live (October 12, 2026).
See what's newResearch agents watch your market, content agents draft posts and comparison pages in your voice, and you choose which outward-facing actions wait for a human yes.
Market brief: what changed since Monday
— Competitor X shipped usage-based pricing [source]
— Two enterprise threads mention agent budgets [source]
— Regulatory draft affects EU deployments [source]
3 claims verified · 1 refuted and omitted
Standing research routines sweep competitor releases, pricing changes, community chatter, and industry news — then deliver a cited brief on your schedule. Ask a one-off question and get a sourced answer, not a search-results paste.
Content agents write blog posts, comparison pages, release notes, and social drafts against your style guide and your actual product. The calendar fills itself; you edit and approve instead of staring at blank pages.
On a new connection, connector actions start as Allowed, including writes. Set the ones that reach the outside world — posts, emails, page edits — to Ask first, and each call waits in Connectors → Review until a person approves it. For work that ships as a task, such as a blog post, add an approval stage so you review the final draft before it’s marked done. Connector settings cover connector calls, not shell commands an agent runs.
Publish “Comparing agent runtimes” to the blog
“Posts on the Paperclip blog are drafted with agents and ship only after a human approves them. That is the whole model: velocity from the machines, taste and accountability from you.”
Market sweeps, drafts in your voice, and a human on publish.