Agent skill

Refresh Agent Skills

by wcygan in wcygan/dotfiles

Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update.

MITAuto-check passedAgent Workflows

Install Refresh Agent Skills

skills CLI
$ npx skills add wcygan/dotfiles --skill refresh-agent-skills -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install wcygan/dotfiles refresh-agent-skills --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/wcygan/dotfiles.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/refresh-agent-skills .claude/skills/refresh-agent-skills && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
refresh-agent-skills
GitHub stars
194
Token cost
~2.4k tokens
SKILL.md length
668 words
Files
1
Skills in repo
19
Repo updated
First seen
Licence
MIT

At a glance

Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update.

  • Works in 7 steps: Inspect state and resolve the provider… → Advance only the consumer pin → Review inventory and stale names → …
  • Asked to update
  • SKILL.md covers Completion contract, Boundaries, Workflow and Quick cheatsheet
  • Calls git, make and gh; reaches github.com

What it does

Refresh Agent Skills is an agent skill from wcygan/dotfiles. Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update. Use when asked to update, refresh, reinstall, reconcile, or publish the project's agent skills.

Its SKILL.md is about 2.4k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows. The repository describes itself as: Configuration for Development Environments. The licence is MIT.

When your agent uses it

  • Asked to update
  • Publish the projects agent skills

Example prompts

  • “/refresh-agent-skills”

Requirements

  • Python 3
  • Docker

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Inspect state and resolve the provider commit
  2. Advance only the consumer pin
  3. Review inventory and stale names
  4. Install through the supported boundary
  5. Verify the installation
  6. Run repository acceptance
  7. Commit and push only in publish mode

What it can do on your machine

Read from SKILL.md and the folder at commit b385ffa. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • make
    • gh
    • python3
    • uv

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • github.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Refresh Agent Skills loads about 2.4k tokens when it runs. Until then it costs about 74 tokens; SKILL.md has 668 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~74
When it runs · the whole SKILL.md, loaded when a task matches
~2.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from wcygan/dotfiles at commit b385ffa, republished under its MIT licence (© wcygan). 668 words, ~2,441 tokens.

Download SKILL.mdSave it as .claude/skills/refresh-agent-skills/SKILL.md (or your agent's skills folder).
name
refresh-agent-skills
description
Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update. Use when asked to update, refresh, reinstall, reconcile, or publish the project's agent skills.
license
MIT
metadata.author
wcygan
metadata.version
0.1.0

Refresh Agent Skills

Update the repository's pinned reusable-skill catalog and reconcile the local Codex installation. Keep wcygan/agent-skills as the source of truth and use the dotfiles bootstrap command as the only consumer-side installer.

Completion contract

The task is complete only when all of these are true:

  • agent-skills.lock.toml contains the reviewed full 40-character lowercase commit and portable shared-user directory for wcygan/agent-skills.
  • ./bootstrap.sh agent-skills succeeds and reports the installed count and destination.
  • ./bootstrap.sh agent-skills --check and make agent-skills-check pass.
  • gh skill list --dir "$HOME/.agents/skills" reports one source, scope=custom, pinned=true, the exact locked version, and one shared user destination for every provider skill.
  • No provider skill disappeared without a stale-local review.
  • make test-pre, git diff --check, and the tracked-lockfile check pass.
  • Commit and push happen only when the invocation explicitly asks to publish.

Boundaries

  • Work from /Users/wcygan/Development/dotfiles (or the current checkout containing bootstrap.sh and agent-skills.lock.toml).
  • Read the nearest AGENTS.md, .agents/skills/dotfiles-operations, and .agents/skills/agent-skills-integration before changing state.
  • Change only the consumer integration and its focused tests and documentation. Do not copy provider files into dotfiles.
  • Preserve unrelated worktree changes. Stage named files only.
  • Keep installed skills and ~/.agents/.skill-lock.json machine-local.
  • Do not run the full dotfiles install, alter Nix inputs, change shell startup files, prune stale skills, or publish a branch unless separately requested.

Workflow

1. Inspect state and resolve the provider commit

Start with repository and provider status. The provider checkout is useful for review, but it is not a second source or installation implementation.

bash
git status --short
git -C ../agent-skills status --short 2>/dev/null || true
sed -n '1,40p' agent-skills.lock.toml
gh skill list --dir "$HOME/.agents/skills" \
  --json skillName,sourceURL,scope,version,pinned,path

Resolve origin/main from the live GitHub remote on every run. Do not use the local ../agent-skills checkout, its current branch, or the existing lock as the latest-version oracle:

bash
remote_refs="$(git ls-remote \
  https://github.com/wcygan/agent-skills.git refs/heads/main)" || exit
if [ "$(printf '%s\n' "$remote_refs" | awk 'NF == 2 { count++ } END { print count + 0 }')" -ne 1 ]; then
  printf '%s\n' "expected exactly one origin/main ref" >&2
  exit 1
fi
latest_sha="$(printf '%s\n' "$remote_refs" | awk 'NR == 1 { print $1 }')"
if ! printf '%s\n' "$latest_sha" | grep -Eq '^[0-9a-f]{40}$'; then
  printf '%s\n' "origin/main did not return a full lowercase SHA" >&2
  exit 1
fi
printf '%s\n' "$latest_sha"

Require exactly one 40-character lowercase hexadecimal SHA from that command. If the remote cannot be resolved, or returns no or multiple refs, stop and report the network/authentication failure. Do not guess a tag, branch, or abbreviated SHA. If the locked commit equals latest_sha, keep the lock unchanged and continue with install/reconcile. Otherwise replace the lock with latest_sha before discovering or installing the catalog.

Review the provider delta before pinning when a provider checkout is present:

bash
git -C ../agent-skills diff --stat \
  "$(sed -n 's/^commit = "\([0-9a-f]\{40\}\)"$/\1/p' agent-skills.lock.toml)..$latest_sha"
git -C ../agent-skills log --oneline --decorate \
  "$(sed -n 's/^commit = "\([0-9a-f]\{40\}\)"$/\1/p' agent-skills.lock.toml)..$latest_sha"
2. Advance only the consumer pin

Use apply_patch to replace only the commit = "..." line in agent-skills.lock.toml. If tests/test_agent_skills.py contains a hard-coded fixture commit, update that fixture to the same full SHA; do not alter the test's behavior or other files.

Then inspect the diff:

bash
git diff -- agent-skills.lock.toml tests/test_agent_skills.py
git diff --check
Show full SKILL.md (281 more words)Show less
3. Review inventory and stale names

The bootstrap installer rejects duplicate or foreign-source collisions. Before using it, inspect the current inventory and, when possible, compare it with the provider tree:

bash
gh skill list --dir "$HOME/.agents/skills" \
  --json skillName,sourceURL,scope,version,pinned,path

find ../agent-skills/skills -mindepth 2 -maxdepth 2 -name SKILL.md \
  -print 2>/dev/null | sort

gh skill install does not prove that removed upstream skills are pruned. If the provider tree is available, identify installed names absent from that tree and report them. Do not delete them automatically; removal is a separate user-authorized cleanup.

4. Install through the supported boundary

Run the focused, journaled mutation only:

bash
./bootstrap.sh agent-skills

Record the reported catalog count and the actual destination. The expected command uses --dir "$HOME/.agents/skills" --all --pin <full-sha> internally; do not replace it with a clone, copy, symlink, or direct write to the shared directory.

5. Verify the installation

Run both the strict read-only check and the Make target:

bash
./bootstrap.sh agent-skills --check
make agent-skills-check

Run this normalized audit so timestamps in GitHub CLI tracking state cannot mask an inventory or metadata problem:

bash
python3 - <<'PY'
import json
import subprocess
from pathlib import Path
import tomllib

with open("agent-skills.lock.toml", "rb") as f:
    lock = tomllib.load(f)
inventory = json.loads(subprocess.check_output([
    "gh", "skill", "list", "--dir",
    str(Path.home() / lock["directory"]),
    "--json", "skillName,sourceURL,scope,version,pinned,path",
], text=True))

names = [item["skillName"] for item in inventory]
assert len(names) == len(set(names)), "duplicate installed skill names"
assert {item["sourceURL"] for item in inventory} == {
    f"https://github.com/{lock['repository']}"
}
assert {item["scope"] for item in inventory} == {"custom"}
assert {item["version"] for item in inventory} == {lock["commit"]}
assert all(item["pinned"] is True for item in inventory)
parents = {str(Path(item["path"]).resolve().parent) for item in inventory}
assert len(parents) == 1, parents
print(f"verified {len(inventory)} skills at {next(iter(parents))}")
PY

If the provider checkout is available, validate its source as an additional provider-side check:

bash
for skill_dir in ../agent-skills/skills/*; do
  [ -f "$skill_dir/SKILL.md" ] || continue
  uv tool run --from skills-ref agentskills validate "$skill_dir" || exit 1
done
gh skill publish --dry-run

Warnings from the provider's dry run are reportable; validation errors are blocking.

6. Run repository acceptance
bash
make test-pre
git diff --check
git diff --exit-code -- flake.lock uv.lock
git status --short
./bootstrap.sh recover

recover must report that no interrupted dotfiles operation needs recovery. Linux, Docker, or evaluation-only results do not establish macOS acceptance; record the actual platform used.

7. Commit and push only in publish mode

When the user explicitly requested commit/push, first confirm the diff contains only the lock/test expectation and this skill. Stage paths explicitly:

bash
git diff --name-only
git add agent-skills.lock.toml src/dotfiles_setup/agent_skills.py \
  src/dotfiles_setup/cli.py tests/test_agent_skills.py Makefile \
  .agents/skills/agent-skills-integration/SKILL.md \
  .agents/skills/dotfiles-operations/references/{architecture.md,operations.md,migrations.md} \
  .agents/skills/refresh-agent-skills/SKILL.md
git diff --cached --check
git diff --cached --stat
git commit -m "chore(skills): refresh pinned agent skills"
git push origin "$(git branch --show-current)"

If the pin was unchanged and the skill file is the only change, use a concise skill-specific commit subject instead. Do not create an empty commit. After a successful push, report the commit SHA, branch, provider SHA, catalog count, destination, checks, stale-name review, and any provider warnings.

Quick cheatsheet

bash
# Resolve the live provider origin/main on every refresh
remote_refs="$(git ls-remote https://github.com/wcygan/agent-skills.git \
  refs/heads/main)" || exit
test "$(printf '%s\n' "$remote_refs" | awk 'NF == 2 { count++ } END { print count + 0 }')" -eq 1
latest_sha="$(printf '%s\n' "$remote_refs" | awk 'NR == 1 { print $1 }')"
printf '%s\n' "$latest_sha" | grep -Eq '^[0-9a-f]{40}$' || exit 1

# Update only agent-skills.lock.toml's commit line with apply_patch

# Reconcile the pinned Codex catalog
./bootstrap.sh agent-skills

# Read-only verification
./bootstrap.sh agent-skills --check
make agent-skills-check

# Full repository gate
make test-pre
git diff --check
git diff --exit-code -- flake.lock uv.lock

# Publish only when explicitly requested
git add agent-skills.lock.toml src/dotfiles_setup/agent_skills.py \
  src/dotfiles_setup/cli.py tests/test_agent_skills.py Makefile \
  .agents/skills/agent-skills-integration/SKILL.md \
  .agents/skills/dotfiles-operations/references/{architecture.md,operations.md,migrations.md} \
  .agents/skills/refresh-agent-skills/SKILL.md
git commit -m "chore(skills): refresh pinned agent skills"
git push origin "$(git branch --show-current)"

© wcygan, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/refresh-agent-skills of wcygan/dotfiles.

Open the folder on GitHubat commit b385ffa

Compare with similar skills

Refresh Agent Skills next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Refresh Agent Skills compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Refresh Agent Skills this skillwcygan/dotfiles194—~2.4kAutomated safety check: PassMIT
MCP Server Builderanthropics/skills180k63 repos~2.3kAutomated safety check: PassApache-2.0
Hook Development for Claude Code Pluginsanthropics/claude-plugins-official38k10 repos~4.1kAutomated safety check: NotesApache-2.0
Using Superpowersfarm-fe/farm5.6k35 repos~1.4kAutomated safety check: PassMIT
Executing Plans Inlineobra/superpowers297k2 repos~5.1kAutomated safety check: PassMIT
Skill CreatorAzure/azqr79689 repos~8.2kAutomated safety check: PassApache-2.0

Similar skills

  • MCP Server Builder

    anthropics/skills

    Official

    Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.

    180k GitHub starsUsed in 63 repos~2.3k tokens
    Agent WorkflowsAuto-check passed
  • Hook Development for Claude Code Plugins

    anthropics/claude-plugins-official

    Official

    Explains how to write Claude Code plugin hooks, both prompt-based checks and bash commands, for events such as PreToolUse, Stop and SessionStart.

    38k GitHub starsUsed in 10 repos~4.1k tokens
    Agent WorkflowsAuto-check: notes
  • Using Superpowers

    farm-fe/farm

    A skill your agent uses when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions

    5.6k GitHub starsUsed in 35 repos~1.4k tokens
    Agent WorkflowsAuto-check passed
  • Executing Plans Inline

    obra/superpowers

    Has the agent carry out an implementation plan itself, task by task in the current session, keeping a ledger, proving each step with a test and ending with one whole-branch review.

    297k GitHub starsUsed in 2 repos~5.1k tokens
    Agent WorkflowsAuto-check passed
  • Skill Creator

    Azure/azqr

    Official

    Create new skills, modify and improve existing skills, and measure skill performance.

    796 GitHub starsUsed in 89 repos~8.2k tokens
    Agent WorkflowsAuto-check passed
  • Claude Code Agent Development

    anthropics/claude-plugins-official

    Official

    Explains how to write agents for Claude Code plugins: the markdown file with YAML frontmatter, trigger descriptions, model and color settings, and system prompt design.

    38k GitHub starsUsed in 7 repos~2.8k tokens
    Agent WorkflowsAuto-check passed

More from wcygan/dotfiles

All 19 skills in this repo
  • Fish Shell Config

    wcygan/dotfiles

    Expert Fish shell configuration including config files, functions, abbreviations, prompts, environment variables, and scripting.

    194 GitHub stars~2.9k tokensUpdated 2 days ago
    Auto-check passed
  • Lazygit Config

    wcygan/dotfiles

    Configure lazygit TUI including config.yml setup, custom pagers (especially delta integration), themes, keybindings, and custom commands.

    194 GitHub stars~3.6k tokensUpdated 2 days ago
    Auto-check passed
  • Nix Manager

    wcygan/dotfiles

    Manage Nix package, profile, flake, and diagnostic work in this dotfiles repository.

    194 GitHub stars~1k tokensUpdated 2 days ago
    Auto-check passed
  • Nvim Config

    wcygan/dotfiles

    Configure Neovim editor with init.lua, plugin management, LSP setup, keymaps, and options.

    194 GitHub stars~4.4k tokensUpdated 2 days ago
    Auto-check passed
  • Starship Config

    wcygan/dotfiles

    Configure and optimize Starship cross-shell prompt with presets, custom modules, and performance tuning.

    194 GitHub stars~1.5k tokensUpdated 2 days ago
    Auto-check passed
  • Dotfiles Operations

    wcygan/dotfiles

    Operate, maintain, upgrade, troubleshoot, and review this dotfiles repository.

    194 GitHub stars~1.4k tokensUpdated 2 days ago
    Auto-check passed

Categories

Questions about Refresh Agent Skills

What does Refresh Agent Skills do?

Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update. Refresh Agent Skills is an agent skill from wcygan/dotfiles. Refresh the pinned wcygan/agent-skills catalog for this dotfiles checkout, install it into Codex user scope, verify every skill and path, and optionally commit and push the update.

When should I use Refresh Agent Skills?

Refresh Agent Skills fits situations like: asked to update; publish the projects agent skills.

How do I install Refresh Agent Skills in Claude Code?

Run `npx skills add wcygan/dotfiles --skill refresh-agent-skills -a claude-code`. Or copy the skill folder (.agents/skills/refresh-agent-skills in wcygan/dotfiles) into .claude/skills/refresh-agent-skills in your project. Claude Code loads it when a task matches its description.

How do I install Refresh Agent Skills in Codex?

Run `npx skills add wcygan/dotfiles --skill refresh-agent-skills -a codex`. Or copy the skill folder (.agents/skills/refresh-agent-skills in wcygan/dotfiles) into .agents/skills/refresh-agent-skills in your project. Codex loads it when a task matches its description.

Can I use Refresh Agent Skills in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add wcygan/dotfiles --skill refresh-agent-skills -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/refresh-agent-skills, .gemini/skills/refresh-agent-skills, .github/skills/refresh-agent-skills and .opencode/skills/refresh-agent-skills in your project.

What does Refresh Agent Skills need to run?

Going by SKILL.md and its folder, Refresh Agent Skills needs the command-line tools its instructions call (git, make, gh, python3 and uv). Our summary lists: Python 3; Docker.

Does Refresh Agent Skills access the network?

SKILL.md names 1 domain. In commands or code: github.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.

Is Refresh Agent Skills safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Refresh Agent Skills use?

Refresh Agent Skills is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Refresh Agent Skills use?

About 2.4k tokens (SKILL.md is roughly 9.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Refresh Agent Skills?

Skills that share tags, products or a category with Refresh Agent Skills: MCP Server Builder (anthropics/skills, 180k stars), Hook Development for Claude Code Plugins (anthropics/claude-plugins-official, 38k stars), Using Superpowers (farm-fe/farm, 5.6k stars) and Executing Plans Inline (obra/superpowers, 297k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Refresh Agent Skills?

wcygan (a GitHub user) maintains it in wcygan/dotfiles, which has 194 GitHub stars. The repository holds 19 skills in this directory. The repository was last updated on October 8, 2026.

Source: wcygan/dotfiles on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.