Neovim Config
Skill0 starsChanged 4 months agoMITView on GitHub
Set up, repair, upgrade, and optimize a Neovim configuration. Use for GREENFIELD setup of a new config tailored to what the user wants, for REPAIRing a broken config (startup errors, failed plugins, LSP/treesitter issues), for UPGRADEing plugins and tooling safely, or for OPTIMIZing an existing working config (startup time, lazy-loading, deprecations, redundancy). lazy.nvim-first; every change is verified by driving headless Neovim.
- Installs packages
In this repository12 other files it gives its agents.
- Brew DoctorSkill
- Credential StoreSkill
- Dev EnvironmentSkill
- DotfilesSkill
- Ghostty ConfigSkill
- Git SetupSkill
- Identity ProfilesSkill
- macOS DefaultsSkill
- Runtime VersionsSkill
- Shell SyncSkill
- SSH ConfigSkill
- Terminal ThemeSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 25
If intent is ambiguous, run Environment discovery below, then confirm which workflow.
- Distinctive: line 59
Verify with headless Neovim: Never declare a change done from reading code — config bugs (wrong API, load order, deprecations) only surface …
- Distinctive: line 68
Don't assume inventory: Read plugin specs from disk each session — they drift.
- Very specific: line 74
Read ~/.config/devenv/config.toml before prompting for choices.
- Distinctive: line 86
Prefer the narrowest change that fixes the problem.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Neovim config toolkit
A generalized toolkit for managing any user's Neovim configuration through four workflows. Works against the real environment on disk — never assume the setup; discover it first.
1. Route to a workflow
The user may pass an argument (setup | repair | upgrade | optimize) or describe the goal
in their own words. Match intent to a workflow, then follow that file:
| The user wants to… | Workflow |
| Create a config (none yet, or a fresh tailored one) | reference/setup.md |
| Fix something broken (errors, a plugin/LSP/treesitter failing) | reference/repair.md |
| Update plugins/tooling and stay current | reference/upgrade.md |
| Review a working config and make it better | reference/optimize.md |
If intent is ambiguous, run Environment discovery below, then confirm which workflow.
2. The scripts (prefer these over ad-hoc bash)
Stable, tested helpers live in scripts/ — call them instead of
re-deriving the commands each session. Run by absolute path from this skill's directory. They
resolve stdpath internally, so never hardcode ~/.config/nvim.
| Need | Script |
| Environment discovery (run first) | scripts/nvim_env.py |
| A plugin's version, pins, tags, drift; locate a symbol across installed/newest-tag/core | scripts/plugin_info.py <plugin> [symbol] |
Runtime verification: startup · open <ext> · ts <ext> · lsp <ext> · deprecations · api <vim.path> | scripts/nvim_check.py <check> … |
| Run arbitrary Lua headlessly, correctly (one-off checks) | scripts/nvim_lua.py '<lua>' |
Full contract + a worked example: scripts/README.md. Reach for nvim_lua.py
for novel checks and the .md snippets below for what the scripts run under the hood.
3. Environment discovery (always run first)
scripts/nvim_env.py prints all of this as key<TAB>value lines. What each fact means:
- Config dir — honor
stdpath('config')($XDG_CONFIG_HOME/nvim,~/.config/nvim, or~/AppData/Local/nvim). Never hardcode~/.config/nvim. - Plugin manager — lazy.nvim if there's a
lazy-lock.jsonin the config dir or alazy/dir understdpath('data'). Also recognize packer, vim-plug, mini.deps, rocks.nvim. This toolkit fully drives lazy.nvim; for other managers, detect and advise, don't automate. - Structure — single
init.luavs a modularlua/tree. Read the real plugin specs to learn the user's conventions before editing. - Data/state — plugins & treesitter parsers live under
stdpath('data'); logs (lsp.log,conform.log) understdpath('state').
4. Core principles (every workflow)
- Verify with headless Neovim. Never declare a change done from reading code — config bugs
(wrong API, load order, deprecations) only surface at runtime. Drive nvim and inspect the
result:
scripts/nvim_check.pyfor the common checks,scripts/nvim_lua.pyfor novel ones; snippets in reference/verification.md. Do a before/after comparison for any fix or optimization. - Check the version first. Package managers auto-upgrade Neovim; a minor bump (e.g. 0.11→0.12) is the single most common cause of sudden breakage.
- Back up before risky changes. Copy
lazy-lock.jsonbefore updates. If the config is a git repo, offer to commit first. If not, keep edits surgical and reversible, and say so. - Don't assume inventory. Read plugin specs from disk each session — they drift.
- Cross-reference known breakages: reference/known-issues.md (Neovim-version bumps, treesitter master→main, mason-lspconfig v2, API deprecations, conform).
5. Settings
Read ~/.config/devenv/config.toml before prompting for choices; if a [nvim-config] section
exists, use it as defaults (precedence: explicit answer this session > config.toml > ask). Had to
ask? Offer to save the answer back. Managed by the devenv skill. Keys honored:
plugin_manager— expected manager (e.g.lazy). Most nvim state is discovered from disk, so this is the main knob; missing → detect it.
6. Safety
- Confirm before deleting or overwriting user config. On setup over an existing config, back it up (timestamped) first — never clobber silently.
- External CLIs must exist before use — verify, don't assume:
git, a C compiler,tree-sitter(vianpm i -g tree-sitter-cli),node/npm,ripgrep, and any formatters/LSP servers. - Prefer the narrowest change that fixes the problem; don't break working LSP/completion chasing a cosmetic cleanup.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

