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 wangke19/gemini-ai-helpers --skill prow-job-analyze-metal-install-failuregit clone --depth 1 https://github.com/wangke19/gemini-ai-helpersWrote 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/wangke19/gemini-ai-helpers/prow-job-analyze-metal-install-failure)<a href="https://agentmods.dev/skills/wangke19/gemini-ai-helpers/prow-job-analyze-metal-install-failure"><img src="https://agentmods.dev/badge/skills/wangke19/gemini-ai-helpers/prow-job-analyze-metal-install-failure/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/wangke19/gemini-ai-helpers/prow-job-analyze-metal-install-failure"><img src="https://agentmods.dev/badge/skills/wangke19/gemini-ai-helpers/prow-job-analyze-metal-install-failure.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 | $0.00070 | $0.05360 |
| Opus 5 | $0.00035 | $0.02680 |
| Sonnet 5 | $0.00014 | $0.01072 |
| Haiku 4.5 | $0.00007 | $0.00536 |
Grade A, and why
Prow Job Analyze Metal Install Failure 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 12d 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 — 460 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prow Job Analyze Metal Install Failure
This skill helps debug OpenShift bare metal installation failures in CI jobs by analyzing dev-scripts logs, libvirt console logs, sosreports, and other metal-specific artifacts.
When to Use This Skill
Use this skill when:
- A bare metal CI job fails with "install should succeed" test failure
- The job name contains "metal" or "baremetal"
- You need to debug Metal3/Ironic provisioning issues
- You need to analyze dev-scripts setup failures
This skill is invoked by the main prow-job-analyze-install-failure skill when it detects a metal job.
Metal Installation Overview
Metal IPI jobs use dev-scripts (https://github.com/openshift-metal3/dev-scripts) with Metal3 and Ironic to install OpenShift:
- dev-scripts: Framework for setting up and installing OpenShift on bare metal
- Metal3: Kubernetes-native interface to Ironic
- Ironic: Bare metal provisioning service
The installation process has multiple layers:
- dev-scripts setup: Configures hypervisor, sets up Ironic/Metal3, builds installer
- Ironic provisioning: Provisions bare metal nodes (or VMs acting as bare metal)
- OpenShift installation: Standard installer runs on provisioned nodes
Failures can occur at any layer, so analysis must check all of them.
Network Architecture (CRITICAL for Understanding IPv6/Disconnected Jobs)
IMPORTANT: The term "disconnected" refers to the cluster nodes, NOT the hypervisor.
Hypervisor (dev-scripts host)
- HAS full internet access
- Downloads packages, container images, and dependencies from the public internet
- Runs dev-scripts Ansible playbooks that download tools (Go, installer, etc.)
- Hosts a local mirror registry to serve the cluster
Cluster VMs/Nodes
- Run in a private IPv6-only network (when IP_STACK=v6)
- NO direct internet access (truly disconnected)
- Pull container images from the hypervisor's local mirror registry
- Access to hypervisor services only (registry, DNS, etc.)
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.
- 12d ago First seen · 460 lines · 70 tokens per session scan A 8e42f8f404e9
Prow Job Analyze Metal Install Failure is a skill published in the GitHub repository wangke19/gemini-ai-helpers (2 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 70 tokens to every session and 5,360 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 2026-08-31.
Other skills, from other repositories
gcloud-usage
This skill should be used when user asks about "GCloud logs", "Cloud Logging queries", "Google Cloud metrics", "GCP observability", "trace analysis", or "debugging production issues on GCP".
spk-admin-setup-doctor
Install, verify, and repair Spec Kitty commands, skills, agent paths, runtime prerequisites, and common setup failures.
lov-fix-codex-session
A recovery tool for stuck Codex conversations, including broken tool-call records and references to a model provider that no longer exists. Codex is an AI coding assistant, and a provider is the service supplying its model.
lov-fix-deepseek-tool-call-error
A recovery and investigation tool for a Codex conversation using DeepSeek, an AI service, when a failed tool request returns “No tool output found for tool call.” It examines local session records and reports how to resume the work.
lov-fix-by-add-log
A debugging assistant that adds a small amount of targeted diagnostic logging and uses the resulting output to investigate a hard-to-reproduce problem.
lov-fix-general
A software troubleshooting workflow for finding the cause of an error, applying a focused fix, and checking the original failing path again.