Prompt Fault Injection — Development skill for Claude Code
Prompt: Fault injection — fire the alert set against real broken states.
How to install Prompt Fault Injection
Installs to ~/.claude/skills/chrisadkin8-k8s-ai-observability-prompt-fault-injection/SKILL.md
mkdir -p ~/.claude/skills/chrisadkin8-k8s-ai-observability-prompt-fault-injection && curl -fsSL https://raw.githubusercontent.com/ChrisAdkin8/k8s-ai-observability/HEAD/prompts/prompt-fault-injection.md -o ~/.claude/skills/chrisadkin8-k8s-ai-observability-prompt-fault-injection/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Prompt Fault Injection does
Prompt: Fault injection — fire the alert set against real broken states
Role & Objective
You are a Kubernetes platform engineer working in the `k8s-ai-observability` repo, implementing **ROADMAP.md item 1**. Twelve pieces of work — one configuration surface, the roadmap's seven modes, the harness they share, and **three faults in the observability path that the roadmap's list does not contain** (finding 6):
- W0 — **one JSON config surface that works unchanged on Kubernetes and Dock
Alternatives in Development
- Metrics Dashboard — Define product metrics dashboard with sources and alert thresholds 7.8k ★
- Fix Imports — Fix Broken Imports 2.8k ★
- Deep Research From Scratch — Deep research has broken out as one of the most popular agent applications 768 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Prompt Injection Triage
Assess whether flagged I/O text is a real prompt-injection / jailbreak that warrants action.
Prompt Injection Bench
Hand-authored indirect prompt-injection attack corpus + honest ASR/FPR benchmark of 4 defenses on a real tool-
Prompt Injection Benchmark
A reproducible prompt-injection benchmark that measures which defenses actually work: each payload is replayed
Diagnose Before Acting
Under operator pressure ("X is broken — fix it"), the first move is to diagnose what's actually broken vs what
Plan Doctor
Self-check for the planning-with-files mechanisms that fail silently: plan resolution, hook injection, canonic
Ac Validation
Review Acceptance Criteria for one or more user stories against completeness and quality standards. Checks hap
Related Agents
AI Code Security Auditor
Security reviewer for AI-generated and vibe-coded apps — hunts the hardcoded secrets, broken row-level securit
Security AI Generated Code Auditor
Security reviewer for AI-generated and vibe-coded apps — hunts the hardcoded secrets, broken row-level securit
Timps API Security Tester
Run an OWASP API Security Top-10 (2023) audit against an OpenAPI spec or live endpoint — broken auth, BOLA, ma