skills/ghfs/SKILL.md
Manages ghfs local mirror files in `.ghfs/`, especially translating user instructions into valid execute operations (`execute.md`, `execute.yml`, or per-item edits), running `ghfs execute` / `ghfs sync`, and validating issue/PR batch edits. Use when tasks involve editing issues/PRs through `.ghfs` artifacts, reconciling sync state, or applying queued GitHub operations.
npx skillsauth add antfu/ghfs ghfsInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use this skill to operate ghfs as a local filesystem mirror for GitHub issues and pull requests.
ghfs sync mirrors remote content into .ghfs/.ghfs execute merges operations from .ghfs/execute.yml, .ghfs/execute.md, and per-item markdown frontmatter differences.--run to mutate GitHub.Key .ghfs files:
execute.md: queued commands (human-friendly)execute.yml: queued operations (YAML array)schema/execute.schema.json: schema for execute.yml.sync.json: sync and execution run history (skip reading it)issues.md, pulls.md, repo.json: aggregated mirror viewsissues/**/*.md, pulls/**/*.md: per-item mirrorsghfs sync..ghfs/execute.md for quick command-style edits (recommended)..ghfs/execute.yml for explicit structured operations..ghfs/issues/**/*.md or .ghfs/pulls/**/*.md frontmatter for direct per-item edits.ghfs execute.ghfs execute --run.Execution behavior to remember:
execute.yml -> execute.md -> per-item generated operations.--run, successful operations are removed from queued YAML/MD sources; failed and not-yet-run entries remain..ghfs/execute.md QuicklyUse one command per line:
close #123 #124
set-title #125 "Improve sync summary output"
label #125 enhancement, cli
assign #126 octocat
comment #126 "Need more context"
close-comment #127 "Closing as completed"
Rules:
# and // line comments plus <!-- ... --> HTML comment blocks are supported and preserved.close, reopen, clear-milestone, unlock, mark-ready-for-review, convert-to-draft.Common aliases:
open -> reopencloses -> closeclose-comment / comment-close / close-and-comment / comment-and-close -> close-with-commentlabel / labels / tag / tags / add-tag -> add-labelsassign / assignee / assignees -> add-assigneestitle / retitle -> set-titlecomment -> add-commentready / undraft -> mark-ready-for-reviewdraft -> convert-to-draft.ghfs/execute.yml PreciselyKeep root as a YAML array and include number + action for each entry.
# yaml-language-server: $schema=./schema/execute.schema.json
- number: 125
action: set-title
title: Improve sync summary output
- number: 125
action: add-labels
labels: [enhancement, cli]
- number: 126
action: request-reviewers
reviewers: [octocat]
ifUnchangedSince: '2026-03-05T04:10:00Z'
- number: 130
action: add-reaction
reaction: heart
- number: 130
action: add-reaction
reaction: '+1'
target: {kind: comment, commentId: 9876543}
Map user intent to action fields:
| User intent | action | Required extra fields |
| --- | --- | --- |
| Close / reopen | close, reopen | none |
| Change title | set-title | title |
| Replace body | set-body | body |
| Add comment | add-comment | body |
| Close with comment | close-with-comment | body |
| Add/remove/set labels | add-labels, remove-labels, set-labels | labels (non-empty string array) |
| Add/remove/set assignees | add-assignees, remove-assignees, set-assignees | assignees (non-empty string array) |
| Set/clear milestone | set-milestone, clear-milestone | milestone for set |
| Lock/unlock conversation | lock, unlock | optional reason for lock |
| PR reviewer actions | request-reviewers, remove-reviewers | reviewers (non-empty string array) |
| PR draft state | mark-ready-for-review, convert-to-draft | none |
| Add/remove a reaction | add-reaction, remove-reaction | reaction (one of +1, -1, laugh, hooray, confused, heart, rocket, eyes); optional target (defaults to the issue/PR body) |
Rules:
number must be a positive integer.ifUnchangedSince must be ISO datetime when present.request-reviewers, remove-reviewers, mark-ready-for-review, and convert-to-draft are PR-only.Practical number resolution:
.ghfs/issues/00123-foo.md -> number: 123..ghfs/issues.md / .ghfs/pulls.md when matching by title.Edit frontmatter in .ghfs/issues/**/*.md or .ghfs/pulls/**/*.md:
titlestate (open / closed)labelsassigneesmilestoneghfs execute compares edited frontmatter against tracked state and generates operations automatically (with ifUnchangedSince safeguards).
Preferred commands:
ghfs sync
ghfs sync --full
ghfs sync --since 2026-03-01T00:00:00Z
ghfs execute
ghfs execute --run
Useful flags:
--repo owner/name when repo cannot be auto-resolved.--non-interactive for scripted runs.--continue-on-error to keep applying later ops after a failure.testing
Create, edit, improve, or audit AgentSkills. Use when creating a new skill from scratch or when asked to improve, review, audit, tidy up, or clean up an existing skill or SKILL.md file. Also use when editing or restructuring a skill directory (moving files to references/ or scripts/, removing stale content, validating against the AgentSkills spec). Triggers on phrases like "create a skill", "author a skill", "tidy up a skill", "improve this skill", "review the skill", "clean up the skill", "audit the skill".
testing
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).
testing
Create, edit, improve, or audit AgentSkills. Use when creating a new skill from scratch or when asked to improve, review, audit, tidy up, or clean up an existing skill or SKILL.md file. Also use when editing or restructuring a skill directory (moving files to references/ or scripts/, removing stale content, validating against the AgentSkills spec). Triggers on phrases like "create a skill", "author a skill", "tidy up a skill", "improve this skill", "review the skill", "clean up the skill", "audit the skill".
testing
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).