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 netlify/axis --skill configure-axisgit clone --depth 1 https://github.com/netlify/axisWrote 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/netlify/axis/configure-axis)<a href="https://agentmods.dev/skills/netlify/axis/configure-axis"><img src="https://agentmods.dev/badge/skills/netlify/axis/configure-axis/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/netlify/axis/configure-axis"><img src="https://agentmods.dev/badge/skills/netlify/axis/configure-axis.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 | $0.00069 | $0.07451 |
| Opus 5.5 | $0.00028 | $0.02980 |
| Sonnet 5.5 | $0.00014 | $0.01490 |
| Haiku 4.5 | $0.00007 | $0.00745 |
Grade C, and why
configure-axis scanned grade C with 2 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 28d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
"teardown": [{ "action": "run_script", "command": "rm -rf node_modules" }], Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsSL https://raw.githubusercontent.com/netlify/axis/main/skills/configure-axis/SKILL.md \ How it starts
The opening of the file, as written. The whole thing — 559 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Configure AXIS
AXIS (Agent Experience Index Score) is a synthetic testing framework for AI agents. This skill teaches you to author the two files an AXIS user maintains:
- Scenarios under
scenarios/(or any path the config points at): one JSON file per task the agent will be asked to perform. axis.config.jsonat the project root: which agents to run, where scenarios live, and what is shared across them.
When to use this skill
Trigger phrases include "set up AXIS", "add an AXIS scenario", "write an axis.config.json", "evaluate my agent with AXIS", "score my agent on X".
Before authoring, do this:
- Look for an existing
axis.config.json(oraxis.config.{js,ts}) at the project root. If one exists, read it; do not overwrite without confirmation. - List the existing scenarios directory if present. Match its naming and style.
- If no config exists, suggest running
npx @netlify/axis initfirst, or scaffold one yourself using the patterns below.
Conceptual model
For each scenario, AXIS runs every configured agent against the same prompt in an isolated workspace, then scores each run on four dimensions and produces an HTML + JSON report.
- Goal achievement (default weight 0.4): did the agent satisfy the judge's checks?
- Environment (0.2): did filesystem / shell / network operations succeed reliably?
- Service (0.2): did external services (APIs, MCP servers) respond reliably?
- Agent (0.2): were the agent's decisions sound across every tool call?
Refer to the framework's output as the AXIS Result. The acronym expands to Agent Experience Index Score.
Authoring a scenario
A scenario is a JSON file under the scenarios directory. The file path (without extension, relative to that directory) becomes the scenario's key. Only name, prompt, and judge are required.
The annotated examples below are labeled
jsoncfor documentation only. They contain// commentsand trailing commas. Real.jsonfiles do NOT support either. When copying these examples into output, strip every//comment line and every trailing comma. Plain JSON examples (in the Recipes section) are safe to copy verbatim.
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.
- 28d ago Changed · +58 lines a72571651637
- 1mo ago Changed · +119 lines · +17 tokens per session 4909ec2d9362
- 2mo ago First seen · 382 lines · 52 tokens per session scan C 5c5d34c45a0b
configure-axis is a skill published in the GitHub repository netlify/axis (66 stars, last pushed yesterday), licensed MIT. It adds 69 tokens to every session and 7,451 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it C with 2 findings (recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
rewardkit
Write Harbor task verifiers using Reward Kit. Use when creating or editing a task's tests/ directory, adding grading criteria, setting up LLM/agent judges, or designing verifiers that produce a reward score.
create-adapter
Scaffold a new Harbor benchmark adapter by running harbor adapter init and then guide implementation using the Adapters Agent Guide as the authoritative spec.
agent-eval
Use when measuring whether an LLM or agent system actually got better and gating merges on it: golden sets, fixing an inflated LLM-as-judge, scoring RAG (faithfulness, contextual recall) or agent trajectories (tool correctness, completion), or picking an eval framework. NOT building the agent loop, tools or RAG…
evaluate-skill
Run, read, and diagnose a skill's Caliper eval — its success rate over k attempts, whether it fires, and whether it beats the agent without it. Use when the user wants to run, validate, interpret, or compare a skill's eval, or write an .eval.yaml spec whose tasks they have already decided.
writing-skills
Use when creating a new skill, editing an existing skill, or verifying skills work before deployment.
evals-implement
Use when a published goldset is ready to become executable graders and configs — Python graders and metrics with auto-run unit tests verifying grader correctness.