parallel-slices

parallel-slices is a skill for Claude Code from paulinevos/claude-stuff. It costs 55 tokens per session (564 once invoked), scanned A, original, MIT.

An orchestrator for splitting repository work among several agents using Jujutsu workspaces. Each slice gets its own directory and dependencies so agents can work in parallel without sharing one working copy.

In plain words
What is it for?
Use it to detect the repository setup, divide a task into dependent slices, create workspaces, and coordinate workers across a shared codebase.
Why use it?
It reduces clashes between simultaneous changes and makes dependencies between pieces of work explicit. It can fall back to Git worktrees when Jujutsu cannot be used.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the jj-workflow plugin — 4 skills shipped together

Good fit Use it to detect the repository setup, divide a task into dependent slices, create workspaces, and coordinate workers across a shared codebase.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/paulinevos/claude-stuff/parallel-slices
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add paulinevos/claude-stuff --skill parallel-slices
Clone the repo
git clone --depth 1 https://github.com/paulinevos/claude-stuff

Made for: Claude Code.

Or install jj-workflow, the plugin that ships this one along with the rest of its 4 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for parallel-slices

README.md
[![agentmods](https://agentmods.dev/badge/skills/paulinevos/claude-stuff/parallel-slices/github.svg)](https://agentmods.dev/skills/paulinevos/claude-stuff/parallel-slices)
Your own site
<a href="https://agentmods.dev/skills/paulinevos/claude-stuff/parallel-slices"><img src="https://agentmods.dev/badge/skills/paulinevos/claude-stuff/parallel-slices/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for parallel-slices

Your own site · 80×15
<a href="https://agentmods.dev/skills/paulinevos/claude-stuff/parallel-slices"><img src="https://agentmods.dev/badge/skills/paulinevos/claude-stuff/parallel-slices.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 55 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 564 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00055 $0.00564
Opus 5.5 $0.00022 $0.00226
Sonnet 5.5 $0.00011 $0.00113
Haiku 4.5 $0.00006 $0.00056

Measured 21d ago against content hash 2c7bc36a4ac6, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

parallel-slices scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 21d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/jj-workflow/skills/parallel-slices/SKILL.md · 46 lines

What it actually says

parallel-slices

You orchestrate; workers own their workspaces. Read conventions. Use git-fallback if jj root fails, or if required work/tests need git submodules. A jj repo with irrelevant submodules may continue, but the workspaces cannot materialise them. Stop if neither jj nor git applies.

  1. Split work into independently testable, reviewable slices. Record name, base, owned files, and done criteria; show the table before creating anything. Independent slices base on trunk(); dependent slices base on the parent's bookmark. Avoid overlapping files.

  2. Before creating a workspace, ask the user to grant edit permission for ${HOME}/.jj-workspaces and wait for that permission. Request the shared root, not each slice directory, so workers are not prompted repeatedly. Then create each workspace outside checkouts, always at ${HOME}/.jj-workspaces/<repo>/<slice>:

    jj git fetch
    repo=$(basename "$(jj root)")
    workspace_root="${HOME}/.jj-workspaces/${repo}"
    mkdir -p "$workspace_root"
    jj workspace add --name <slice> -r 'trunk()' --sparse-patterns full "$workspace_root/<slice>"
    

    For a dependent slice, use -r <parent-slice> only after that bookmark exists. Do not create a bookmark on its initial empty @.

  3. Give each worker its workspace path, slice/bookmark and base, purpose, owned files, hand-off criteria, and work-in-slice. They must not touch other slices. If a child starts before hand-off, its parent may add at the tip but must not rewrite ancestors.

  4. Monitor from the default workspace with revsets, not commands in worker directories. A plain jj -R <workspace> snapshots that worker's edits; if -R is unavoidable, add --ignore-working-copy.

  5. After hand-off, use finish-slices. Rebase a child onto a grown parent only from the default workspace; its now-stale handed-off workspace is harmless.

All workspaces share revisions and bookmarks immediately. One owner per change, bookmark, and workspace prevents concurrent rewrites from making workers stale.

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 21d ago Changed · -60 lines · -130 tokens per session 2c7bc36a4ac6
  2. 1mo ago First seen · 106 lines · 185 tokens per session scan A c82df31f56e6

Subscribe to this mod's changes

parallel-slices is a skill published in the GitHub repository paulinevos/claude-stuff (64 stars, last pushed 4d ago), licensed MIT. It adds 55 tokens to every session and 564 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-10.

Related

Other skills, from other repositories