What it does
Turns what your team already answers into content they stop having to write. It mines resolved tickets for questions asked repeatedly, checks whether the knowledge base already covers them, and drafts the missing article from the answers your own agents actually gave — not from a generic template. The same pass produces macros. Where several agents answered the same question in slightly different words, that is a macro waiting to exist; the agent drafts one, notes which resolutions it was built from, and flags where the existing answers disagreed with each other. Disagreement between agents is usually the more valuable finding — it means the policy itself is unclear. Everything is a draft. Articles are created unpublished and macros are created inactive, so a human reviews before anything reaches a customer or a colleague's sidebar. Nothing is ever published, activated, or edited over. It refuses to write what it cannot ground. If the resolved tickets do not contain a real answer — only escalations, or replies that promised a follow-up — it says the question is unanswered rather than inventing a plausible-sounding procedure.
Example prompts
- What did customers ask most this month that we have no article for?
- We answer the SSO timeout question constantly — draft the article from how we've actually been answering it.
- Draft a macro for password reset requests and tell me which tickets you built it from.
Before you connect
The credentials this agent will ask you for — the full setup is on the Setup tab.
Needs 1 credential (1 required) to connect. See setup
