Index of every page on this site: llms.txt
All skills
scode avatar

/skillette

@03a9f7f
Run it yourself

Use when the user invokes `skillette`, `$skillette`, or any `skillette-<name>` trigger.

  • 25 files
  • 237.6 KB
  • Updated last month
  • GitHub
Use this Skill: https://skilld.dev/gh/scode/dotfiles/skillette

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈25 tokens for metadata: the name and description. ≈582 when used: this file. ≈44k more on demand in 20 files.

Description uses 1.2% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈99 characters, or 1.2%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

skillette

On a trigger match, read <name>/SKILLETTE.md in this skill's base directory, where <name> is the row's first trigger minus its skillette- prefix. The whole message is the request. The base directory is the directory containing this SKILL.md; if that path is no longer in context, reload skillette through your harness's skill mechanism (the Skill tool on Claude Code, the skill tool on OpenCode, the read_skill tool on Muse Code; Codex has no loader and the base directory is ${CODEX_HOME:-$HOME/.codex}/skills/skillette). If the loader fails or the file is unreadable, stop and name the path or tool; do not continue from memory or from a search elsewhere. An explicit skillette- word wins over any phrase match; if two rows still plausibly match, ask which; if a skillette- word matches no row, say so. Bare skillette with no trigger just confirms the skill is loaded.

Triggers Natural-language triggers
skillette-brain use or create a brain, its artifacts, or a code review queue; not skillette changes
skillette-change add, remove, or change a skillette, or change skillette itself
skillette-codex-wake-on-exit in Codex: wake up later, check periodically, or wait for a condition; not discussion or edits
skillette-cr-triage invoking cr-triage; not skillette changes
skillette-hackmd the user mentions HackMD or hackmd.io
skillette-lore the whole word "lore" meaning repo history or a ./lore dir
skillette-ntfy a request to send via ntfy or $ntfy, including after other work
skillette-scripts about to write a script or non-trivial inline code in any language
skillette-snippets the user mentions scode-snippets, a scode snippet, or "the preview flow"

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 03a9f7f. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated last month

README badge

README badge for scode/dotfiles/skillette