Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add kazdenc/builder-skills --skill metricsgit clone --depth 1 https://github.com/kazdenc/builder-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kazdenc/builder-skills/metrics)<a href="https://agentmods.dev/skills/kazdenc/builder-skills/metrics"><img src="https://agentmods.dev/badge/skills/kazdenc/builder-skills/metrics/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kazdenc/builder-skills/metrics"><img src="https://agentmods.dev/badge/skills/kazdenc/builder-skills/metrics.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00065 | $0.01591 |
| Opus 5 | $0.00032 | $0.00796 |
| Sonnet 5 | $0.00013 | $0.00318 |
| Haiku 4.5 | $0.00006 | $0.00159 |
Grade A, and why
metrics scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Define what success looks like in numbers. Work through each step — don't jump to a list of metrics without grounding them in the goal first.
Step 1: Identify the Goal
Before picking metrics, answer:
- What job does this feature serve? What progress is the user trying to make? (Reference the jtbd skill if available.)
- What outcome matters to the business? Revenue, retention, efficiency, expansion — pick one.
- What behavior change are you expecting? Users should do more of X, less of Y, start doing Z.
If you can't state the goal in one sentence, the metrics will be unfocused. Push for clarity.
Step 2: Define a Primary Metric
Pick the ONE number that best represents success. This is the metric the team checks every morning.
How to choose — the 3 A's:
| Test | Question | Bad example | Good example |
|---|---|---|---|
| Actionable | Can the team change this number with their work? | Total registered users (too broad) | Weekly active users of this feature |
| Accessible | Can everyone on the team understand and check it? | Custom composite score | % of users who complete core action |
| Auditable | Can you verify the data is correct? | Self-reported satisfaction | Event-tracked completion rate |
Common primary metrics by goal:
| Goal | Primary metric |
|---|---|
| Adoption | % of eligible users who use the feature in first 7 days |
| Engagement | Weekly active usage (sessions, actions, or time) |
| Retention | % still using after 30 days |
| Efficiency | Time to complete task (before vs. after) |
| Revenue | Conversion rate or revenue per user |
| Quality | Error rate or task success rate |
State it precisely: "% of users who create at least one invoice within 7 days of first seeing the feature" — not "adoption."
Step 3: Define Supporting Metrics
Leading indicators
Predict future success. Move before the primary metric does. Use these to course-correct early.
| Example | What it predicts |
|---|---|
| Activation rate (completed setup) | Future retention |
| Feature discovery rate (saw the entry point) | Future adoption |
| Time to first value | Future engagement |
| Onboarding completion | Future active usage |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 158 lines · 65 tokens per session scan A c3dd72dcfd35
metrics is a skill published in the GitHub repository kazdenc/builder-skills (45 stars, last pushed 6mo ago), licensed MIT. It adds 65 tokens to every session and 1,591 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
commercial-opportunity-review
Review Product, offering, catalog, quote, and sale evidence for commercial opportunities.
demand-triage
Prepare classification, evidence gaps, and score inputs for product demand without crossing the funding boundary.
investment-preparation
Prepare an evidence-linked product investment packet while preserving the governed funding decision.
outcome-review
Review whether a funded product bet changed its intended outcome and append evidence-backed learning.
product-line-performance-review
Roll real Product evidence into a ProductLine review without changing the canonical reporting boundary.
stakeholder-brief
Prepare a concise owner or stakeholder brief with current decisions, evidence, bets, outcomes, roadmap, and provenance.