Sign inGet started
Skills

Project knowledge

View as MarkdownOpen in ChatGPTOpen in Claude
Connect your agent
Add to CursorAdd to VS Code

The team's instructions, goals and memory your agent reads before every answer and every run, who writes them, the limits and how agents add to them.

Updated

Project knowledge is what your agent knows about the business before it looks at a number. get_context returns it with every question and every run, so a line written here is a line nobody repeats in every chat.

What does it hold?

PartWho writes itExample
Team instructionsOwners and admins, up to 20,000 characters"Answer in Ukrainian. Keep it to five lines."
Your own instructionsEach person, up to 2,000 characters"Show money in EUR."
Memory: About the businessAnyone, up to 200 characters each"We close on Fridays."
Memory: DecisionsAnyone"We stopped TikTok ads in October."
Memory: What workedAnyone"Shorter pricing page raised trials by 12%."

Edit it in Project knowledge in the dashboard. Owners and admins make the agent forget a memory.

How do agents add to it?

  • remember saves a memory after you agree: "Remember that our busy season is December."
  • add_note puts a dated event on the timeline: "Note that we raised prices on November 3."
  • A scheduled run's suggest_memory leaves what it learned.

Each memory reaches your agent with who wrote it. Your agent reads a teammate's memory as information, never as an instruction: it never outranks what you ask, and it never makes your agent change conversions or forget a visitor without your yes.

What does my agent know?

Ask it. This prompt shows what it knows and what's missing:

Read get_context for shop.example and tell me what you know about the business, what's missing, and three facts I should add.

Next step

Steer skills

Related
  • How a run works: A scheduled run, step by step: it starts, reads context and numbers, reports a finding or stays quiet and finishes.
  • Anatomy of a finding: One finding from start to fix, told as a story: our rule notices wasted ad spend, a skill runs and explains it, and a person decides what to change.
  • Steer skills: Shape what skills look at and how they report, with team instructions, a skill's own text and feedback.
  • Built-in skills: The built-in analytics skills every nextanalytica project starts with: what each one does, the schedule it gets when added and what it needs.
  • What's in a skill: A skill's parts: a name, when to use it and the instructions, as one SKILL.md you can write yourself.
  • Insights: Signals from our rules and findings from your skills, newest first, with the numbers that prove them.
Still have questions?
Was this page useful?