Skip to content

ericrisco/rsc-harness / skills/kpi-framework

kpi-framework

Verified

Use when a team must decide what to measure before building anything — picking one north-star metric, separating leading input drivers from lagging outputs, adding guardrails so a number cannot be gamed, and setting a target that is not arbitrary. NOT the live dashboard that displays them (that is `dashboard`), NOT instrumenting the events (that is `analytics`), NOT the recurring board report (that is `reporting`).

FileKindSize
SKILL.mdskill md11.5 KB
evals/README.mdmarkdown777 B
evals/cases.yamlconfig3.1 KB
references/definition-and-targets.mdmarkdown3.4 KB
references/metric-catalog.mdmarkdown4.6 KB

evals/cases.yaml

View on GitHub
skill: kpi-framework

should_trigger:
  - prompt: "What should we even be measuring for our new product?"
    why: No agreed metric exists yet — the core 'choose what to measure' trigger.
  - prompt: "We track 40 KPIs and none of them lead anywhere — help us pick what actually matters."
    why: KPI sprawl with no leading metric; needs one north star plus a driver set. Non-obvious framing — the symptom is too many metrics, not too few.
  - prompt: "Is 'signups' a leading or lagging indicator for us?"
    why: Leading-vs-lagging classification of an output metric, the input/output distinction at the heart of the skill.
  - prompt: "Set an activation-rate target that won't get gamed."
    why: Non-obvious — pairs target-setting with Goodhart's Law and guardrail design, not just a number.
  - prompt: "quins KPIs triem per al nostre producte SaaS?"
    why: Catalan trigger — 'which KPIs do we pick for our SaaS product', the metric-selection decision.
  - prompt: "Audit our dashboard metrics for vanity numbers and tell us which ones are un-actionable."
    why: Vanity-metric and actionability audit of existing KPIs — the reject/classify side of the skill.

should_not_trigger:
  - prompt: "Build the dashboard that displays our north-star metric."
    route_to: dashboard
    why: Displaying a chosen metric is a display/build task; kpi-framework only decides which metric.
  - prompt: "Instrument the activation event in our app and set up the funnel."
    route_to: analytics
    why: Event instrumentation and funnels are analytics implementation, downstream of the metric decision.
  - prompt: "Write the monthly metrics report for the board."
    route_to: reporting
    why: Producing the recurring narrative around numbers is reporting, not choosing the numbers.
  - prompt: "Design an A/B test to lift our activation rate."
    route_to: ab-testing
    why: Designing an experiment to move an input is ab-testing; kpi-framework supplies the input to move.
  - prompt: "Project our MRR forward 12 months given current growth."
    route_to: forecasting
    why: Projecting a metric over time is forecasting, not defining which metric matters.

capability:
  - scenario: "A seed-stage B2B SaaS team says 'we want more usage' and asks what they should measure. They have signups and total logins tracked, nothing else. Produce a metric definition doc."
    must_include:
      - One output north-star metric tied to retained/delivered value, expressed as a rate or ratio (not a raw count like total signups or total logins).
      - 3-5 leading, directly controllable input drivers stated as concrete instrumentable events (e.g. % of new accounts reaching an activation milestone in week 1), not abstractions like 'engagement'.
      - At least one guardrail/countermetric with the specific gaming risk it protects against named.
      - A target for the north star with an explicit baseline, a calibrated magnitude (not 'double it'), a target date, and an owner.
      - An explicit rejection of a vanity metric (e.g. total signups or total logins) with the reasoning for why it is rejected.
      - A handoff note that instrumentation goes to analytics and display goes to dashboard.
deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall