1. Write a skill
An admin opens Settings → AI context and creates a skill with a name and a few lines of instructions — how a metric is defined, when to use a specific chart type, or which fields matter.
Skills
Teach Basedash once.
It learns everywhere.
Write your team's metric definitions, conventions, and playbooks as reusable skills. Every Basedash AI surface — chat, charts, dashboards, automations, insights — picks them up automatically.
14-day trial. No credit card required.
Settings / Organization / AI context
AI context
Configure AI behavior for your organization. This context is shared with all members.
Global context
Always loaded by every AI surface.
We're a SaaS company. Users are called "merchants".
MRR excludes trial accounts. Revenue is GAAP, not bookings.
Skills
5
Reusable instructions agents load when relevant.
New skillActivation rate
Active in first 7 days · excludes trial accounts · calendar weeks (UTC)
Updated 2d ago
Revenue analysis
Segment by plan tier · exclude trials from MRR · prefer line charts for trends
Updated 2d ago
Cohort definitions
Cohorts are signup_month · retention is plan-active in week N
Updated 2d ago
Support triage
Group tickets by category + priority · open > 48h is SLA risk
Updated 2d ago
Experiments
A/B reports use exposure_first events · 7-day windows · show p-values
Updated 2d ago
How skills work
Define a concept. Apply it everywhere.
Skills are short, plain-language playbooks for a single concept. Write one, and every AI agent in Basedash can pick it up when it's relevant.
2. AI reads it on demand
When a teammate asks a relevant question, the agent loads the full skill before answering. You'll see the read step right in the thinking trace, so the source of every answer is auditable.
3. Every surface uses it
The same skill catalogue feeds chat, the chart builder, dashboards, automations, insights, and background tasks. One definition, every agent.
A skill, up close
Plain language. No SQL required.
A skill is a name and a few lines of instructions. Write it the way you'd brief a new analyst — define the concept, list the rules, and call out any pitfalls. Up to 50,000 characters per skill, so even your most nuanced playbooks fit.
- Semantic layer — define how MRR, activation, churn, and retention should be calculated.
- Team conventions — finance uses GAAP revenue, growth uses calendar weeks, support uses SLA windows.
- Analytical playbooks — how to write A/B reports, how to triage tickets, how to summarize cohorts.
- Chart preferences — "use a line chart for activation trends, a bar chart for plan breakdowns."
AI context / Activation rate
Edit skill
Activation rate
Name
Activation rate
Instructions
Describe when and how AI should use this skill.
- 1 Activation = signed up AND completed onboarding within 7 days.
- 2 Exclude trial-only accounts and internal email domains.
- 3 Use calendar weeks in UTC; weeks start on Monday.
- 4 Always compare cohorts by signup_month, not by event_month.
- 5 Prefer a line chart for the trend over time.
Skill updated · 2 minutes ago
Save changes
Skills work everywhere
One catalogue. Every agent.
Skills aren't tied to a single feature. The same definitions power every AI surface in Basedash, so your team's rules apply whether someone is asking a quick question or running a scheduled report.
Chat
Answer questions using your team's metric definitions.
Chart builder
Build charts that follow your conventions.
Dashboards
Lay out dashboards using your terminology.
Automations
Run scheduled reports using the same rules.
Insights
Surface trends that match your definitions.
Tasks
Plan and complete background work with shared context.
What a skill looks like
Three skills, three teams.
Skills work best when each one focuses on a single concept or workflow. Here's the kind of thing real teams write.
Activation rate
Semantic layer for growth metrics
- Activation = signed up and completed onboarding within 7 days.
- Exclude trial-only accounts and internal email domains.
- Use calendar weeks in UTC; weeks start on Monday.
Revenue analysis
Finance team's standard playbook
- MRR is GAAP recurring revenue, excluding trial accounts.
- Always segment revenue by plan tier in any breakdown.
- When comparing periods, use calendar months in UTC.
Support triage
Operational rules for the support team
- Group tickets by category and priority before summarizing.
- Tickets open longer than 48 hours are SLA risk.
- Use human-readable status names, not raw enum values.
Frequently asked
Questions about skills.
What are skills in Basedash?
Skills are reusable bundles of instructions that you write once and every Basedash AI surface can read on demand. Each skill captures a single concept — a metric definition, a team playbook, an analytical convention — so the AI applies your team's rules without you re-explaining them in every prompt.
Where do skills apply?
Skills are picked up across chat, the chart builder, dashboards, automations, insights, and background tasks. Any time an AI agent in Basedash takes on a request, it sees the catalogue of skills your organization has defined and loads the ones that look relevant before answering.
How does the AI choose which skill to load?
Every agent gets a lightweight catalogue of skill names in its system prompt. When a skill looks relevant to the current request, the agent fetches its full instructions through a tool call you can see in the thinking trace — so there is no black-box retrieval, and you can audit which skills were used.
Who can create or edit skills?
Skills are managed by organization admins under Settings → AI context. Anyone in your workspace benefits from them automatically, but only admins can create, edit, or delete them — so the shared context stays consistent across every team that uses Basedash.
How are skills different from global context?
Global context is a single block of instructions that is always loaded by every AI agent. Skills are modular: many of them, each focused on a single concept, loaded on demand only when relevant. Global context is best for short, durable facts. Skills are best for richer playbooks like metric definitions, team conventions, and analytical workflows.
Can skills act as a semantic layer?
Yes. The most common use is exactly that: defining how the AI should calculate key business metrics — activation, MRR, churn, retention — so every chart, report, and answer treats those concepts the same way across the company.
Get started in under 30 minutes
We can help you migrate your data and dashboards from any other tool.