Agent skill

Sweep Workspace

by fastrepl in fastrepl/anarlog

Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace.

MITAuto-check passed

Install Sweep Workspace

skills CLI
$ npx skills add fastrepl/anarlog --skill sweep-workspace -a claude-code

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

GitHub CLI
$ gh skill install fastrepl/anarlog sweep-workspace --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/fastrepl/anarlog.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/sweep-workspace .claude/skills/sweep-workspace && 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
sweep-workspace
GitHub stars
9.4k
Token cost
~2k tokens
SKILL.md length
1,102 words
Files
1
Skills in repo
32
Repo updated
First seen
Licence
MIT

At a glance

Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace.

  • Works in 3 steps: Merge PRs that are already ready → Update the workspace → Run the focused workflows in order
  • SKILL.md covers Applicability, 1. Merge PRs that are already…, 2. Update the workspace and 3. Run the focused workflows…, plus 1 more section
  • Calls git

What it does

Sweep Workspace is an agent skill from fastrepl/anarlog. Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace. Coordinate a GitButler workspace PR sweep by merging already-ready PRs, updating the workspace, then running create-prs, fix-prs, and update-prs in that order.

Its SKILL.md is about 2k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: Open source Granola AI Alternative. The licence is MIT.

Example prompts

  • “/sweep-workspace”

Workflow steps

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

  1. Merge PRs that are already ready
  2. Update the workspace
  3. Run the focused workflows in order

What it can do on your machine

Read from SKILL.md and the folder at commit 259a04e. 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

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

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    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

Sweep Workspace loads about 2k tokens when it runs. Until then it costs about 69 tokens; SKILL.md has 1,102 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~69
When it runs · the whole SKILL.md, loaded when a task matches
~2k

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 fastrepl/anarlog at commit 259a04e, republished under its MIT licence (© fastrepl). 1,102 words, ~2,013 tokens.

Download SKILL.mdSave it as .claude/skills/sweep-workspace/SKILL.md (or your agent's skills folder).
name
sweep-workspace
description
Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace. Coordinate a GitButler workspace PR sweep by merging already-ready PRs, updating the workspace, then running create-prs, fix-prs, and update-prs in that order.

Sweep GitButler workspace

Applicability

Use only in the Anarlog repository or its checkouts/worktrees when the active checkout branch is exactly gitbutler/workspace. Before running this workflow, check with git symbolic-ref --quiet --short HEAD (a read-only branch check). On any other branch or detached HEAD, skip this skill and use the normal Git workflow. Do not initialize GitButler or switch branches to activate the skill. Use this checkout's skills, not a global copy or another checkout's skills.

Use this skill when asked to sweep PRs in the Anarlog GitButler workspace.

Read the repository-local but skill and the active repository's AGENTS.md and workflow skill when present. Use but for version control and the repository's preferred authenticated forge tools for PR operations. Derive the repository and configured target from the current workspace.

This skill coordinates the three focused workflows. Read the repository-local skill when its stage begins: create-prs, fix-prs, and update-prs. Do not automatically spawn subagents or run stages concurrently.

A sweep invocation authorizes in-scope PR creation, pushes, necessary history rewrites, workspace updates, conflict repairs, review replies/resolutions, metadata edits, and ready non-draft merges. Preserve unrelated work and active collaborators' edits. Editing this skill does not execute a sweep.

1. Merge PRs that are already ready

  • Inspect but status --json and map applied branches to live open PRs by exact head repository/branch. Record bases, remote SHAs and stack dependencies. Honor a narrower user scope; skip missing, closed or merged PRs in this stage.
  • Inspect current published code, all paginated reviews/threads/conversation findings and current checks. Merge only non-drafts with passing applicable required checks, satisfied review requirements, no conflicts, and no remaining actionable or unverified findings. Include outdated unresolved findings and linked reviewer reports; incomplete access/coverage blocks readiness.
  • Do not fix code, publish local changes or create PRs to make this initial queue ready. An unpublished fix does not clear a finding on the remote head. A fix in an upper PR qualifies only when that verified fix is included in the same ready merge prefix; an earlier prefix without it is not safe.
  • Merge nearest trunk first, in batches of at most 10 PRs total. Prerequisites must already be merged or included in the same supported stack merge. Skip blocked ancestors and their descendants; continue independent ready stacks.
  • Immediately before each request, recheck every included PR's head, base, draft state, checks, reviews and mergeability. Use the repository's merge method and the reviewed SHA; changed heads invalidate readiness.
  • Prefer a supported GitButler PR merge operation with head protection. Otherwise use authenticated forge tools: ordinary GitHub PRs support gh pr merge <PR> --match-head-commit <SHA> with the repo's merge method. Native GitHub stacks need the supported asynchronous stack merge API; verify its current request fields and head guard rather than guessing or falling back to a legacy merge. Neither but land nor enabling auto-merge substitutes for this stage.
  • Confirm each requested merge through live terminal PR/operation state before advancing dependent PRs or cleaning up branches. Re-read affected descendants after each confirmed merge. Async acceptance is not merge completion; retain pending operation IDs for continuation and do not mutate affected stacks while their merge is pending.

2. Update the workspace

Before any workspace update, confirm pending merge operations have reached a terminal state. While a merge is pending, defer but pull and all create, fix, and metadata-update operations on its affected stacks. Continue independent stacks only when the available operations can demonstrably leave pending stacks untouched; but pull --check is inspection, not permission to rebase them. If pending stacks cannot be excluded from the workspace update, defer the update and dependent stages until confirmation, retaining operation IDs and reviewed heads for the follow-up.

After this guard is satisfied, update from the configured target with but pull so confirmed merges are incorporated before running the focused workflows. With other agents' branches applied, use but pull --check first to understand affected work. The sweep authorizes scoped updates; work outside that scope needs authorization before it moves. Preserve dirty work and follow the but skill's parking/restoration rules only for in-scope changes. Do not repoint the target.

Record any conflicts introduced by the update for fix-prs. If those conflicts prevent PR creation, let create-prs report the blocked branches, then resolve them in the next stage; retry creation for those branches on the next sweep. If the update itself cannot complete, report its blocker instead of claiming the workspace is current or proceeding on dependent work.

Show full SKILL.md (371 more words)Show less

3. Run the focused workflows in order

  1. create-prs — create missing PRs for the remaining scoped branches.
  2. fix-prs — resolve workspace conflicts and fix active PR reviews and CI.
  3. update-prs — rewrite active PR titles/descriptions from their complete published changes, including the repairs just made.

Pass the repository, scope, current stack map, known blockers, and pending merge operation IDs, affected branch names and reviewed heads between stages. Exclude stacks with pending merges from every stage; child workflows must not pull, publish, edit metadata, or otherwise mutate those stacks. Refresh live branch/PR mappings after mutations. A blocked branch or pending CI does not prevent independent work in later stages. Run the metadata stage after all currently actionable fixes are prepared/published, even when remote checks are pending. Local-only changes remain an explicit description gap.

Do not merge newly created or repaired PRs at the end of this pass. Once ready, they enter the initial merge stage of the next sweep.

Continue when external work is pending

The coordinator owns follow-up scheduling; child workflows report waits to it instead of creating duplicate monitors. Finish actionable stages before waiting. Use available per-PR review/CI subscriptions, or reuse/create a thread heartbeat every 10 minutes when subscriptions are unavailable. Do not watch, sleep or repeatedly poll in the foreground. Schedule a next sweep for newly created/repaired PRs awaiting the initial merge stage even if their checks are already green; use the thread heartbeat when no further subscription event is expected. Retain already-ready PRs deferred solely by the merge batch limit as actionable work and schedule their next sweep even when no external checks are pending.

Save repo, scope, stack map, remote heads, findings/coverage, pending merge IDs, completed stages, conflicts and required user decisions. Resume by confirming pending merges, then run the same merge → workspace update → create → fix → update sequence using fresh evidence. A timer does not supply user approval. Notify only for meaningful changes, completion, failure or required action. Pause when all scoped work is complete, or only user decisions/drafts remain and no actionable work or external checks are pending. Report scheduling failures.

Report confirmed merges, workspace update, PRs created, repairs, metadata updates and remaining blockers briefly. Keep local verification, remote checks, merges and deployment as separate facts.

© fastrepl, 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/sweep-workspace of fastrepl/anarlog.

Open the folder on GitHubat commit 259a04e

Compare with similar skills

Sweep Workspace 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.

Sweep Workspace compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Sweep Workspace this skillfastrepl/anarlog9.4k—~2kAutomated safety check: PassMIT
Checkoutremotion-dev/remotion62k—~166Automated safety check: PassCustom licence
Branchesredis/RedisInsight8.9k—~409Automated safety check: PassCustom licence
Conventional Branchgithub/awesome-copilot40k—~1.1kAutomated safety check: PassMIT
Finishing a Development Branchobra/superpowers296k5 repos~1.9kAutomated safety check: PassMIT
Git Branch Namingmakeplane/plane60k—~594Automated safety check: PassAGPL-3.0

Similar skills

  • Checkout

    remotion-dev/remotion

    Official

    Checkout a requested branch or ref in the Remotion repo, rename the current task to the associated PR title when running in Codex, then install dependencies and build.

    62k GitHub stars~166 tokensUpdated today
    Media & CreativeAuto-check passed
  • Branches

    redis/RedisInsight

    Official

    Create and name git branches following project conventions. An agent skill from redis/RedisInsight.

    8.9k GitHub stars~409 tokensUpdated 3 days ago
    DevelopmentAuto-check passed
  • Conventional Branch

    github/awesome-copilot

    Official

    Create Git branches following the Conventional Branch specification (feature/, bugfix/, hotfix/, release/, chore/).

    40k GitHub stars~1.1k tokensUpdated today
    DevelopmentAuto-check passed
  • Walks the last step of a branch: confirm tests pass, detect the git environment, ask how to integrate, carry out your choice and clean up the worktree.

    296k GitHub starsUsed in 5 repos~1.9k tokens
    DevelopmentAuto-check passed
  • Git Branch Naming

    makeplane/plane

    Names a new Git branch with a type prefix, the lowercased work item ID and a short kebab-case description, so the ID can be extracted later from the branch name.

    60k GitHub stars~594 tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Brooks Sweep

    sickn33/agentic-awesome-skills

    Full-sweep mode: runs a unified analysis across all quality dimensions — code decay, architecture, tech debt, and test quality — then applies fixes directly to the codebase.

    47k GitHub starsUsed in 1 repo~668 tokens
    DevelopmentAuto-check passed

More from fastrepl/anarlog

All 32 skills in this repo
  • But

    fastrepl/anarlog

    Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace.

    9.4k GitHub stars~5.8k tokensUpdated today
    Auto-check passed
  • Create PRs

    fastrepl/anarlog

    Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace.

    9.4k GitHub stars~1.2k tokensUpdated today
    Auto-check passed
  • Fix Ready PRs

    fastrepl/anarlog

    Inspect every open non-draft PR for CI failures and unresolved Cursor Bugbot findings, then fix them on the existing PR branches.

    9.4k GitHub stars~1.4k tokensUpdated today
    Auto-check passed
  • No Use Effect

    fastrepl/anarlog

    Avoid direct React useEffect usage when writing or reviewing React components and hooks.

    9.4k GitHub stars~825 tokensUpdated today
    Auto-check passed
  • QA Critical UX

    fastrepl/anarlog

    QA Anarlog's critical Pro user journey on a signed staging candidate — onboarding, responsive launch, microphone and system-audio capture, automated summaries, and cloud sync.

    9.4k GitHub stars~1.4k tokensUpdated today
    Auto-check passed
  • Reactive Sqlite UI

    fastrepl/anarlog

    Build SQLite-backed reactive UI in apps/desktop using stable patterns for reads, selection, forms, writes, and loading states.

    9.4k GitHub stars~699 tokensUpdated today
    Auto-check passed

Questions about Sweep Workspace

What does Sweep Workspace do?

Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace. Sweep Workspace is an agent skill from fastrepl/anarlog. Use only in the Anarlog repository when the active checkout branch is exactly gitbutler/workspace.

How do I install Sweep Workspace in Claude Code?

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

How do I install Sweep Workspace in Codex?

Run `npx skills add fastrepl/anarlog --skill sweep-workspace -a codex`. Or copy the skill folder (.agents/skills/sweep-workspace in fastrepl/anarlog) into .agents/skills/sweep-workspace in your project. Codex loads it when a task matches its description.

Can I use Sweep Workspace 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 fastrepl/anarlog --skill sweep-workspace -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/sweep-workspace, .gemini/skills/sweep-workspace, .github/skills/sweep-workspace and .opencode/skills/sweep-workspace in your project.

What does Sweep Workspace need to run?

Going by SKILL.md and its folder, Sweep Workspace needs the command-line tools its instructions call (git).

Does Sweep Workspace access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Sweep Workspace 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 Sweep Workspace use?

Sweep Workspace is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Sweep Workspace use?

About 2k tokens (SKILL.md is roughly 8.1k 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 Sweep Workspace?

Skills that share tags, products or a category with Sweep Workspace: Checkout (remotion-dev/remotion, 62k stars), Branches (redis/RedisInsight, 8.9k stars), Conventional Branch (github/awesome-copilot, 40k stars) and Finishing a Development Branch (obra/superpowers, 296k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Sweep Workspace?

fastrepl (a GitHub organization) maintains it in fastrepl/anarlog, which has 9,445 GitHub stars. The repository holds 32 skills in this directory. The repository was last updated on October 7, 2026.

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