pskoett-ai-skills
Self-verifying coding workflows with failure recovery and learning loops
Gives your agent durable working principles for this suite: state assumptions before coding, keep changes minimal and surgical, define verifiable success criteria, and route failures into the self-healing and self-improvement skills.
All Skills in This Repository (9)
Pure Emerald Level Indicatorsintent-framed-agent
Capture coding session intent and monitor for scope drift.
agent-teams-simplify-and-harden
Orchestrate parallel agent teams to implement, audit, and refine code iteratively.
simplify-and-harden
Automate a three-pass code review for simplification, security hardening, and micro-documentation.
self-improvement
Log errors, corrections, and knowledge gaps during AI agent execution.
self-improvement-ci
Capture recurring failure patterns from pull request checks in CI/CD pipelines.
simplify-and-harden-ci
Execute headless CI scans for simplification, hardening, and documentation on pull requests.
context-surfing
Monitor AI context window health and detect drift during multi-step execution.
plan-interview
Gather software requirements through structured interviews before implementation.
dx-data-navigator
Query the DX Data Cloud PostgreSQL database to analyze developer productivity and engineering metrics.
Frequently Asked Questions
FAQPage SchemaHow to install pskoett-ai-skills?▼
Run `npx skills add pskoett/pskoett-ai-skills --all -g -y` in your terminal to install the full skill set globally.
What problem does pskoett-ai-skills solve?▼
It stops coding agents from shipping unverified work, drifting off-scope, and repeating the same mistakes across sessions. Each skill targets one failure mode, from planning interviews to verified failure recovery.
How does the self-healing skill work?▼
When a command, test, or build fails mid-task, the agent diagnoses the root cause, applies a minimal patch, re-runs the failing operation to prove the fix, and files a reusable HEAL entry for future sessions.
Does pskoett-ai-skills work with Claude Code, Codex, and Copilot?▼
Yes. Every skill follows the open Agent Skills specification and ships with plugin manifests for Claude Code, Codex, and GitHub Copilot CLI.
Can I use these skills in CI pipelines?▼
Yes. Headless CI variants like self-healing-ci and simplify-and-harden-ci run inside GitHub Actions to review pull requests and capture failure patterns without interactive prompts.
Related Repositories in Software Engineering
View All in Software Engineering→openclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core