Analysis updated 2026-05-18
Give an AI assistant a strict protocol to inspect existing code before changing it.
Keep an AI assistant from rewriting working code it was not asked to touch.
Carry decisions and context forward across multiple AI coding sessions using a snapshot summary.
Apply a consistent safe-editing discipline to an AI assistant across mobile, web, or backend projects.
| bonushora/surgical-dev-ops | 0-bingwu-0/live-interpreter | 010zx00x1/faresnipe | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | — | Python | Python |
| Setup difficulty | easy | moderate | easy |
| Complexity | — | 2/5 | 2/5 |
| Audience | developer | general | general |
Figures from each repo's GitHub metadata at analysis time.
README is in Portuguese by default, with an English version linked in the repo.
Surgical DevOps is not a piece of software you install and run. It is a set of written rules, called protocols, meant to be given to an AI coding assistant so it behaves more carefully when it changes existing code. The README is written in Portuguese, with an English translation also included in the repository. The project defines two protocols. The first, called BH-SEP, tells an AI to inspect existing code before changing anything, to preserve working code instead of rewriting it, to make the smallest possible changes, and to check its work after each step rather than making a pile of edits at once. The second, called BH-SDP, deals with keeping track of decisions and context across separate coding sessions, by having the AI produce a structured summary, called a snapshot, of what was decided and what files were touched, so the next session can pick up where the last one left off. To use it, you copy a ready made block of instructions into your AI assistant that points it to the two protocol files hosted on GitHub and tells it to follow them strictly for the rest of the session. The assistant is expected to confirm it has understood the rules before continuing. The author describes this as language and framework independent, meant to apply to mobile apps, web apps, backend systems, and infrastructure code alike, wherever AI assistance is used to modify an existing codebase. The stated goal is to reduce bugs and regressions caused by an AI assuming things about a system it has not actually inspected, while keeping the human in charge of engineering decisions.
A set of protocol instructions you paste into an AI coding assistant to make it inspect code carefully, avoid big rewrites, and track context across sessions.
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.