Analysis updated 2026-05-18
Give AI coding agents clear guardrails for scoping, building, and reviewing work.
Coordinate multiple AI agents acting as orchestrators, builders, and reviewers.
Track deployment evidence so agent work is verifiable in production, not just claimed.
| joelbrilliant1-beep/agentic-delivery-skills | 0verflowme/alarm-clock | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | — | CSS | Python |
| Last pushed | — | 2022-10-03 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | easy | easy | moderate |
| Complexity | 2/5 | 2/5 | 4/5 |
| Audience | developer | vibe coder | developer |
Figures from each repo's GitHub metadata at analysis time.
Agentic Delivery Skills is a collection of reusable skill files for teams using AI coding agents, tools like Claude Code, Codex, Cursor-style agents, or custom multi-agent systems, to build and ship software. Each skill is a plain Markdown file containing structured instructions that an agent can load to follow a specific operating pattern rather than improvising its own approach. The problem these skills address is that AI agents doing software work often lack clear role boundaries, leave their work invisible to the humans supervising them, and drift beyond their intended scope. One example given is a reviewer that silently rewrites code instead of simply flagging issues for a human to decide on. These skills exist to provide practical guardrails across the whole delivery pipeline, from planning through review to deployment. The library is organized into two groups. Core delivery skills cover clarifying scope and breaking work into vertical slices, meaning small, independently reviewable pieces of work, coordinating the roles of orchestrators, builders, and reviewers, and applying adversarial code review that goes beyond simply checking whether tests pass. Operational guardrail skills cover adjacent concerns: promoting local work to production with verifiable evidence at each stage, keeping agent work visible and traceable through run IDs, checking that a new AI model or provider actually supports the capabilities a team needs before switching to it, handling long sessions and context handoffs without losing state, and importing external code safely. The skills are designed to be portable across different agent frameworks. Installing them means copying the relevant Markdown files into an agent's own skill directory. The underlying philosophy is clear separation of roles: one builder assigned per slice of work, reviewers who document problems before attempting to fix them, and deployment evidence that proves what is actually running in production, not just what the repository contains.
A library of Markdown skill files that give AI coding agents clear role boundaries and guardrails throughout the software delivery pipeline.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.