How do I install the Cx Programming skill?

Run npx skillmds@latest add bongsuchoi/cx-programming in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Cx Programming safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Cx Programming?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Cx Programming free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Cx Programming?

BongSuCHOI (@bongsuchoi) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Cx Programming by bongsuchoi — Agent Skill

Name: Cx Programming
Author: BongSuCHOI

Cx Programming

Apply non-obvious Python, TypeScript, Go or Rust guidance at typing, validation, async/concurrency, resource, error, FFI or toolchain boundaries. Use when language-specific failure modes matter; do not load for routine syntax or when project conventions already decide the approach.

BongSuCHOI 93e8780 27 files · 23.6 KB Updated Sep 17, 2026

File contents

Programming

Load only the language card needed for the current decision. Project conventions and dependencies win.

Priority

Follow the nearest project AGENTS.md, manifest, architecture and style.
Reuse existing dependencies and patterns.
Load only the reference needed for the current failure mode.
Prefer the smallest implementation that satisfies observable behavior.

Reference routing

Python: references/python/README.md
TypeScript: references/typescript/README.md
Go: references/go/README.md
Rust: references/rust/README.md

Reference recommendations are decision support, not reasons to replace an established stack or upgrade dependencies.

Implementation and verification

Inspect call sites and trust boundaries before changing behavior.
Use strict types where they reduce material invalid states; avoid ceremony that adds no safety.
Add or update tests when behavior changes or regression risk warrants it.
Follow project verification first; otherwise run the closest relevant test, typecheck, lint or build.
Do not repeat equivalent verification.

For behavior-preserving refactors, state the behavior that must remain unchanged, trace affected callers and boundaries, and use characterization coverage only when risk justifies it. Preview syntax-aware bulk rewrites and inspect the resulting diff. Do not combine a refactor with dependency upgrades or unrelated cleanup.

Helper

scripts/python/new-script.py creates a no-overwrite PEP 723 script skeleton. Run it with uv run --isolated. If the default uv cache fails specifically with a sandbox permission error, retry once with UV_CACHE_DIR=/tmp/codex-uv-cache; do not relabel package resolution, network, script, or target failures as cache failures. If uv is missing, state why it is needed and propose the exact installation command. Preserve normal helper failures as execution errors.

Use ast-grep only when syntax-aware matching materially improves on rg. If it is missing, continue safely when practical or propose the exact installation command and wait for approval.

---name: cx-programmingdescription: Apply non-obvious Python, TypeScript, Go or Rust guidance at typing, validation, async/concurrency, resource, error, FFI or toolchain boundaries. Use when language-specific failure modes matter; do not load for routine syntax or when project conventions already decide the approach.---# ProgrammingLoad only the language card needed for the current decision. Project conventions and dependencies win.## Priority1. Follow the nearest project `AGENTS.md`, manifest, architecture and style.2. Reuse existing dependencies and patterns.3. Load only the reference needed for the current failure mode.4. Prefer the smallest implementation that satisfies observable behavior.## Reference routing- Python: `references/python/README.md`- TypeScript: `references/typescript/README.md`- Go: `references/go/README.md`- Rust: `references/rust/README.md`Reference recommendations are decision support, not reasons to replace an established stack or upgrade dependencies.## Implementation and verification- Inspect call sites and trust boundaries before changing behavior.- Use strict types where they reduce material invalid states; avoid ceremony that adds no safety.- Add or update tests when behavior changes or regression risk warrants it.- Follow project verification first; otherwise run the closest relevant test, typecheck, lint or build.- Do not repeat equivalent verification.For behavior-preserving refactors, state the behavior that must remain unchanged, trace affected callers and boundaries, and use characterization coverage only when risk justifies it. Preview syntax-aware bulk rewrites and inspect the resulting diff. Do not combine a refactor with dependency upgrades or unrelated cleanup.## Helper`scripts/python/new-script.py` creates a no-overwrite PEP 723 script skeleton. Run it with `uv run --isolated`. If the default `uv` cache fails specifically with a sandbox permission error, retry once with `UV_CACHE_DIR=/tmp/codex-uv-cache`; do not relabel package resolution, network, script, or target failures as cache failures. If `uv` is missing, state why it is needed and propose the exact installation command. Preserve normal helper failures as execution errors.Use `ast-grep` only when syntax-aware matching materially improves on `rg`. If it is missing, continue safely when practical or propose the exact installation command and wait for approval.

Frequently asked questions