用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/pacifio/atlas --skill atlas-self-configure命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG/vector/code-native assets, extending an established icon or logo system, or building the visual directly in HTML/CSS/canvas.
Create and scaffold plugin directories for Codex with a required `.codex-plugin/plugin.json`, optional plugin folders/files, valid manifest defaults, and personal-marketplace entries by default. Use when Codex needs to create a new personal plugin, add optional plugin structure, generate or update marketplace entries for plugin ordering and availability metadata, or update an existing local plugin during development with the CLI-driven cachebuster and reinstall flow.
Create or update a Codex skill with appropriately scoped instructions and any needed supporting resources.
| name | atlas-self-configure |
| description | Inspect and safely update Atlas user preferences through Atlas's config.toml. |
Atlas's user-facing preferences live in one human-editable file,
config.toml. This skill is how you (an agent) read and change it safely —
without touching anything else Atlas persists.
~/.config/atlas/config.toml
The same path on every platform. If $XDG_CONFIG_HOME is set to an absolute
path, it's $XDG_CONFIG_HOME/atlas/config.toml instead — check that variable
before assuming.
Editing that file is the supported mechanism, and the only one available to you: Atlas's config commands are IPC endpoints reachable from its own UI, not agent tools. Atlas watches the file and validates every change you make (see step 6 below), so a direct edit is a first-class way in — not a workaround.
If the file does not exist yet, Atlas hasn't created it (a fresh install before first launch, or the user deleted it). Do not create one speculatively — report that to the user instead of guessing at defaults.
Atlas writes a comment above every key: what it does, its default, and any constraint on its value — a numeric range, an exact set of allowed strings, "must not be empty". Read those comments and follow them. They are generated from the same table Atlas validates against, so they cannot drift from what it will actually accept.
Do not work from a schema you remember, or from one you saw in another project. The file in front of you is authoritative, and it is the only thing guaranteed to match the Atlas build the user is running.
Two things the comments won't spell out:
[settings] block, not just the lines with
values on them. To clear one of these, delete its line; never write an
empty string.Read the whole file first. You need its comments and formatting to avoid disturbing them, and you need the target key's comment to know what a valid value even is.
Explain what you're about to change to the user before writing: the key, its current value, the new value, and what it does — quoting the file's own comment. Don't silently flip a setting.
Change only the line (or few lines) you mean to change. Leave every other key, comment, and blank line exactly as it was — including the comment above the key you're editing, which is Atlas's to rewrite, not yours. Do not reformat, reorder, or re-indent anything:
# Inline git blame — a dim author/age/summary annotation trailing the
# active line in the editor. (default: true)
-someSetting = true
+someSetting = false
Validate the value against that key's comment before writing. If the comment states a range or a fixed set of values and the user asked for something outside it, say so and stop; don't write it and let Atlas reject it.
Write atomically: write your edited content to a new file in the
same directory (e.g. config.toml.tmp.<random>), then rename/move it
over config.toml. Never edit in place — a crash or a concurrent Atlas
reload mid-write must not leave a half-written file behind.
Re-read the file after writing to confirm your change landed.
Atlas re-validates the whole file whenever it notices a change, live, no
restart. An invalid edit is rejected entirely and Atlas keeps running on the
last good settings — but the file stays as you wrote it, and while it's
invalid Atlas also refuses every settings write from its own UI, until someone
repairs it or uses "Recreate defaults" (which overwrites it, keeping a
.bak-<unix-seconds> copy). So a bad write is not free: report it and offer to
restore the exact content you read in step 1, rather than trying to patch your
way out.
config.toml.state.json, device.json, telemetry.json — separate files with
separate ownership; not in scope for this skill.~/.zshrc and similar)..atlas/ state — none of
that is a "setting."If a request needs any of the above, say so plainly and stop — it's out of scope for this skill, not something to work around.