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

/scode-chores

@03a9f7f
Run it yourself

Run periodic repository maintenance chores and prepare one stacked PR per approved chore. Use when the user explicitly invokes /scode-chores or $scode-chores to check dependency-lock updates, dprint plugin updates, or GitHub Actions version updates.

  • 2 files
  • 6 KB
  • Updated 5 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/scode/dotfiles/scode-chores

Nothing lands on disk. Nothing to clean up.

Fork this Skill

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

SKILL.md

≈66 tokens for metadata: the name and description. ≈1.4k when used: this file. ≈76 more on demand in 1 file.

Description uses 3.3% 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 ≈264 characters, or 3.3%.

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.

Scode Chores

Run routine dependency and tooling update chores, then offer the user the PRs that are worth creating. This is not a general modernization pass. It should only make boring update PRs whose main purpose is keeping existing dependencies, formatting plugins, and GitHub Actions current.

Workflow

  1. Detect which chores apply from the repository root.
  2. For each applicable chore, run the smallest update command that can prove whether a real diff exists.
  3. Record the exact diff summary, then restore the worktree before trying the next chore. Do not leave discovery diffs piled on top of each other.
  4. Present the full option list to the user:
    • applicable chores with an upgrade diff
    • applicable chores that were already current
    • chores that were skipped because the project does not use that tool
    • chores that were blocked because a required command or external lookup failed
  5. Ask which upgrade PRs to create. If the user already said to do all available chores, treat every chore with a real diff as approved.
  6. Use $jjstack for the approved chores. By default, create one PR per chore, in the checklist order below. Do not combine Rust, dprint, and GitHub Actions updates into a single PR unless the user explicitly asks.
  7. Recreate each approved chore's diff in its own stack entry, run the relevant verification for that chore, and let $jjstack create or update the stacked PRs.
  8. Summarize which PRs were created, which chores were already current, which were skipped, and which were blocked.

Before starting discovery, inspect VCS state. If the worktree has unrelated user changes, do not overwrite or revert them. Either use a temporary worktree for discovery or record enough state to revert only the files changed by the current chore.

Checklist

1. Update Rust lockfile dependencies

Detect: The repository has a Cargo.toml at the root. Treat this chore as PR-worthy only when Cargo.lock is tracked or otherwise intentionally committed.

Skip if: There is no Rust project, or the project does not commit a Cargo.lock.

Discovery command:

cargo update

PR condition: Cargo.lock changed.

Why: cargo update keeps the existing semver constraints but refreshes the lockfile to newer compatible crate versions. That makes dependency drift visible in a small, reviewable PR instead of bundling it into unrelated work.

Verify: Run the repository's normal Rust checks when they are documented. If there is no repo-specific guidance, run cargo test. For workspaces where that is too expensive or unavailable, run the narrowest command that proves the updated lockfile still resolves, and report the limitation.

2. Update dprint WASM plugins

Detect: The repository has a root dprint config such as dprint.json, dprint.jsonc, .dprint.json, or .dprint.jsonc.

Skip if: There is no root dprint config.

Discovery command:

dprint config update

If the project uses a different dprint config path, use the project's path instead of assuming the root file.

PR condition: The dprint config changed, normally under the plugins list.

Why: dprint pins formatter plugins by URL. Updating those URLs keeps formatting behavior current while making any formatter-version change explicit.

Verify: Run dprint check. If the plugin update changes formatting behavior, run dprint fmt, include the formatting diff in the same dprint PR, then run dprint check again.

3. Update GitHub Actions versions

Detect: The repository has .github/workflows/*.yml or .github/workflows/*.yaml, and at least one workflow step has a uses: value that points at a GitHub Action.

Skip if: There are no GitHub Actions workflows, or no external actions are referenced.

Discovery process:

  • Find action references with rg 'uses:\s*[^./][^@]+@' .github/workflows.
  • Ignore local actions (./...), Docker image actions (docker://...), and non-version refs that are intentionally branches or SHAs unless the user asks to pin or change them.
  • For each version-tagged action, check the upstream repository's current stable tag using GitHub releases/tags or an available action-update tool. Prefer official GitHub data over third-party indexes.
  • Update to the newest stable version tag. Avoid prereleases unless the existing ref is already a prerelease or the user explicitly asks for prereleases.

PR condition: One or more workflow uses: refs changed.

Why: Actions are executable CI dependencies. Keeping them current picks up security fixes and platform compatibility updates without mixing CI dependency bumps into feature work.

Verify: Read the workflow diff and confirm every changed uses: ref points at an existing upstream tag. Run any local workflow linter the repository already uses. If no local validation exists, report that GitHub Actions validation will happen in CI after the PR is opened.

PR Boundaries

Use these default commit and PR scopes:

  • Rust dependency chore: chore: update rust dependencies
  • dprint plugin chore: chore: update dprint plugins
  • GitHub Actions chore: ci: update github actions

Keep each approved chore as one stack entry even when it changes multiple files. For example, a dprint plugin update and the formatting changes caused by that update belong in the same dprint PR.

If one approved chore depends on another, say so before creating PRs and put the prerequisite lower in the stack. Otherwise use the checklist order.

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 5 months ago

README badge

README badge for scode/dotfiles/scode-chores