generated: '2026-08-13'
method: generated
source: >-
  Packaged Agent Skills generated from openapi/_original/pirsch-pirsch-api-openapi.yml
  and the Pirsch v1 reference at https://docs.pirsch.io/api-sdks/api-v1. Every
  operationId named in a skill was verified verbatim against the spec.
  Cross-cutting rules cite ../conventions/, ../errors/, ../authentication/,
  ../rate-limits/ and ../components/. Pirsch publishes no AGENTS.md or skill
  pack of its own — searched docs.pirsch.io and the pirsch-analytics GitHub
  organization on 2026-08-13.
skills:
  - file: pirsch-report-domain-traffic.md
    name: Report traffic for a Pirsch domain
    api: openapi/pirsch-statistics-api-openapi.yml
    operations: [getToken, listDomains, getTotalStatistics, getVisitorStatistics, getPageStatistics, getReferrerStatistics, getChannelStatistics]
  - file: pirsch-track-server-side.md
    name: Track page views and events from a server
    api: openapi/pirsch-tracking-api-openapi.yml
    operations: [getToken, sendPageView, sendPageViewBatch, sendEvent, sendEventBatch, keepSessionAlive, keepSessionAliveBatch]
  - file: pirsch-configure-goals-and-funnels.md
    name: Configure conversion goals and funnels
    api: openapi/pirsch-conversion-goals-api-openapi.yml
    operations: [listConversionGoals, createConversionGoal, updateConversionGoal, deleteConversionGoal, testGoalRegex, listFunnels, createOrUpdateFunnel, deleteFunnel, getGoalStatistics, getFunnelStatistics]
  - file: pirsch-onboard-a-domain.md
    name: Onboard a new domain end to end
    api: openapi/pirsch-domains-api-openapi.yml
    operations: [createDomain, listDomains, updateDomainSettings, createAlternativeDomain, createClient, inviteMembers, listMembers, updateMember, createAccessLink]
cross_cutting_rules:
  - >-
    No idempotency support anywhere in the API — every skill states the retry
    consequence explicitly rather than assuming safe retries.
  - >-
    Access keys (pa_) are write-only; statistics reads require an OAuth client
    token.
  - >-
    Statistics scope by `id` (domain id); writes scope by `domain_id`. The two
    are not interchangeable.
  - >-
    Errors are a proprietary {validation, error, context} envelope with no
    machine-readable code — branch on message text or not at all.