Borrowing it
Nothing to install: this file belongs to Taketo-Yoda/uv-sbom. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Taketo-Yoda/uv-sbom/develop/.claude/skills/split/SKILL.mdgit clone --depth 1 https://github.com/Taketo-Yoda/uv-sbomWrote 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.
[](https://agentmods.dev/skills/taketo-yoda/uv-sbom/split)<a href="https://agentmods.dev/skills/taketo-yoda/uv-sbom/split"><img src="https://agentmods.dev/badge/skills/taketo-yoda/uv-sbom/split/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.
<a href="https://agentmods.dev/skills/taketo-yoda/uv-sbom/split"><img src="https://agentmods.dev/badge/skills/taketo-yoda/uv-sbom/split.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00032 | $0.02660 |
| Opus 5.5 | $0.00013 | $0.01064 |
| Sonnet 5.5 | $0.00006 | $0.00532 |
| Haiku 4.5 | $0.00003 | $0.00266 |
Grade A, and why
split 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 15d 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.
How it starts
The opening of the file, as written. The whole thing — 272 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/split - Issue Decomposition Skill
Decomposes a large GitHub Issue into subtask Issues, each implementable autonomously in a single PR.
Language Requirement
IMPORTANT: All created GitHub Issues MUST be written in English.
Workflow Overview
Read Parent Issue → Propose Decomposition → Wait for Confirmation → Create Subtask Issues → Comment on Parent
Steps
Step 1: Read the Parent Issue (MANDATORY)
gh issue view <issue-number>
Extract:
- Issue title and overall scope
Acceptance Criteria(human section) andTechnical Acceptance Criteria(AI section)Files to Update/Create(inside the parent Issue's collapsed🤖 Implementation Spec (for AI agents)block —gh issue viewreturns it in the raw body)- Any existing dependencies or constraints
Step 2: Analyze and Propose Decomposition
Identify subtask boundaries using these dimensions:
| Dimension | Example |
|---|---|
| Module boundary | Changes touch independent modules (e.g., adapters/ vs domain/) |
| Concern boundary | "add struct" vs "add tests" vs "update CLI flag" |
| Sequential dependency | Subtask B requires subtask A to merge first |
| Risk boundary | Mechanical rename vs logic change should be separate |
Granularity rule: Each subtask must be implementable in a single focused PR.
Guard rails:
- If the issue is already small enough for one PR → report "No split needed. This issue is already scoped for a single PR." Do NOT create any subtask Issues.
- If decomposition yields more than 7 subtasks → flag over-decomposition and suggest grouping before proceeding.
- Line-count target: Aim for ≤ 200 changed lines per subtask PR. If a subtask is estimated to exceed this, consider splitting it further. This is a guideline, not a hard limit.
Shared-Dependency Convergence Check (conditional)
Trigger: two or more proposed subtasks each change a different call site, and all of them converge on the same shared module, type, or helper — whether that shared item is introduced by the parent Issue, by an earlier sibling subtask, or already exists. "Migrate N call sites onto a new shared helper" is the canonical shape.
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.
- 15d ago Changed · +3 lines 786d3e5ace91
- 16d ago Changed · -2 lines f6bfb9c95a7b
- 18d ago Changed · +14 lines 81bef45acd6c
- 19d ago Changed · +14 lines 38718c917259
- 1mo ago Changed · +73 lines b051f68bb908
- 1mo ago First seen · 170 lines · 32 tokens per session scan A 78b9febc86d3
split is a skill published in the GitHub repository Taketo-Yoda/uv-sbom (6 stars, last pushed yesterday), licensed MIT. It adds 32 tokens to every session and 2,660 once invoked, about $0.0001 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 2026-09-04.
Other skills, from other repositories
select-build-target
Pull the next source contract to build from the pending backlog, respecting outside-in dependency order.
uv-docs
Astral uv 0.11.x — fast Python package and project manager in Rust. Projects, scripts, tools, Python versions, pip interface.
aip-tracker
Track Airflow Improvement Proposal (AIP) implementation progress by comparing Confluence specs against codebase evidence. Use when asked to assess, report on, or compare AIP status.
mempalace-task
Create, hand off, claim, execute, and close agent tasks through the MemPalace logstream. Use when the user wants to delegate work, prepare a ready-to-paste task for another agent, receive a MemPalace task id, or explicitly launch a supported headless agent in controlled mode.
intake
Inspect configured Sources and propose evidence-backed Backlog work.
refresh
Refresh Product status from current work and source evidence.