A pull-request analyzer that counts hand-written lines changed and applies 200-line and 600-line split checks. A pull request is a proposed code change; excluded files include generated, vendor, minified, and lock files.

Use it to analyze a PR or file list, report included and excluded lines, require split analysis at 200 lines, and recommend separate PR slices at 600 lines.

It separates meaningful code churn from files that inflate change counts. It also makes large changes trigger a recorded analysis or multiple smaller PRs.

split-pr — Skill by jl-cmd

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.

agentmods

npx agentmods add skills/jl-cmd/claude-dev-env/split-pr

Any agent

npx skills add jl-cmd/claude-dev-env --skill split-pr

Clone the repo

git clone --depth 1 https://github.com/jl-cmd/claude-dev-env

Made for: Claude Code, Codex.

Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,140 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

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.

Model	Per session	Once invoked
Fable 5	$0.00043	$0.01140
Opus 5	$0.00022	$0.00570
Sonnet 5	$0.00009	$0.00228
Haiku 4.5	$0.00004	$0.00114

Measured 2d ago against content hash 14cd511bc7dd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

split-pr 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 2d ago.

The scan reads SKILL.md. This mod also ships 33 executable files (scripts/analyze_pr.py, scripts/categorize_files.py, scripts/config/__init__.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

packages/claude-dev-env/.agents/skills-archived/split-pr/SKILL.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

split-pr (analyzer)

Source-owned analyzer for the locked hand-written line metric and the 200 / 600 split gates. Reports hand-written churn separately from excluded churn (generated, vendor, minified, lockfile). File count informs judgment and is never a hard gate.

Gotchas

Hand-written lines are additions plus deletions on non-excluded paths.
200 hand-written lines require a recorded split_analysis document.
600 hand-written lines default to multiple PR slices (default_split).
Atomic exception at 600+ needs a recorded unsplittable reason and a standing Fable verdict token — never stay one PR silently.
File count is context only. It never blocks analysis.

When this applies

User asks for a split analysis or hand-written line report on a PR or file list.
A PR approaches or exceeds the 200 / 600 hand-written line surfaces.

Process

Run python scripts/analyze_pr.py --files-json <path> --pretty (or --pr N).
Read hand_written_lines, excluded_churn_lines, requires_split_analysis, default_split, and optional atomic_exception.
At 200+ without a prior analysis artifact, record the analyzer JSON before shipping further commits on that surface.
At 600+ with default_split=true, plan multiple PR slices unless atomic_exception carries a Fable-approved reason.

File index

File	Purpose
`SKILL.md`	Analyzer contract and 200/600 gates
`scripts/analyze_pr.py`	Emit analysis JSON from gh or offline files
`scripts/categorize_files.py`	Path churn class (hand-written vs excluded)
`scripts/split_pr_script_types.py`	Canonical split-plan build and path-assignment validation
`scripts/split_pr_title.py`	Collapse titles to exactly one conventional prefix
`scripts/split_pr_layer_order.py`	Deterministic config→other layer sort
`scripts/split_pr_paginate.py`	Paginated PR file intake via gh api --slurp
`scripts/pack_files_into_slices.py`	Pack annotated files into ≤200 hand-written-line slices by layer
`scripts/config/packing_constants.py`	Review budget, hard cap, path→layer markers for packing
`scripts/test_categorize_files_packing.py`	Budget packing and layer inference tests
`scripts/verify_plan.py`	Coverage verification: unique path assignment, path normalization, title contract
`scripts/test_verify_plan.py`	Full-coverage and gap rejection tests
`scripts/test_verify_plan_contract.py`	source_commit and title-normalization contract tests
`scripts/test_verify_plan_path_normalization.py`	Unsafe-path fail-closed tests
`scripts/split_pr_dependency_graph.py`	Deterministic layer-rank dependency graph over slices
`scripts/config/dependency_constants.py`	Graph key names and layer ranks
`scripts/test_split_pr_dependency_graph.py`	Graph order, duplicate-id, unknown-layer tests
`scripts/verify_dependency_graph.py`	CLI + shape/order checks for a dependency graph document
`scripts/test_verify_dependency_graph.py`	Rebuild-order accept/reject and CLI exit tests
`scripts/config/split_pr_constants.py`	Thresholds and excluded path markers
`scripts/config/plan_constants.py`	Plan schema keys, layer order, title-prefix tokens
`scripts/test_analyze_pr.py`	Boundary tests at 199/200/599/600
`scripts/test_categorize_files.py`	Exclusion classification tests
`scripts/test_split_pr_script_types.py`	Plan schema and full path assignment
`scripts/test_split_pr_title.py`	One-prefix title normalization
`scripts/test_split_pr_layer_order.py`	Layer order ranking
`scripts/test_split_pr_paginate.py`	Slurp page flattening
`scripts/split_pr_process_runner.py`	Captured subprocess runner for Git ops
`scripts/split_pr_git_operations.py`	Safe local slice commit + restore via git_operations_constants
`scripts/execute_split_slices.py`	Materialize full plan in topo order
`scripts/config/git_operations_constants.py`	Git argv tokens and error strings
`scripts/test_split_pr_git_operations.py`	Dirty-tree, path staging, restore tests
`scripts/test_execute_split_slices.py`	Plan walk order and failure restore
`reference/proposal-format.md`	Human-readable plan field contract
`reference/splitting-principles.md`	Vertical-slice invariants
`reference/path-layers.md`	Layer hints (context only)
`reference/split-further-loop.md`	Re-split loop until leaves fit review

Read the full file on GitHub · 86 lines

Files

What ships with it

37 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.

2d ago First seen · 86 lines · 43 tokens per session scan A 14cd511bc7dd

Subscribe to this mod's changes

split-pr is a skill published in the GitHub repository jl-cmd/claude-dev-env (6 stars, last pushed 2d ago), licensed MIT. It adds 43 tokens to every session and 1,140 once invoked, about $0.0002 per session on Opus 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 2026-08-31.

Code review

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

agent-host-chat-contributions

Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.

microsoft/vscode · 56 tokens

auto-perf-optimize

Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.

microsoft/vscode · 62 tokens