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.
npx skills add mostafa-drz/claude-skills --skill chunk-prgit clone --depth 1 https://github.com/mostafa-drz/claude-skillsWrote 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/mostafa-drz/claude-skills/chunk-pr)<a href="https://agentmods.dev/skills/mostafa-drz/claude-skills/chunk-pr"><img src="https://agentmods.dev/badge/skills/mostafa-drz/claude-skills/chunk-pr/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/mostafa-drz/claude-skills/chunk-pr"><img src="https://agentmods.dev/badge/skills/mostafa-drz/claude-skills/chunk-pr.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 | $0.00004 | $0.03967 |
| Opus 5 | $0.00002 | $0.01983 |
| Sonnet 5 | $0.00001 | $0.00793 |
| Haiku 4.5 | $0.00000 | $0.00397 |
Grade A, and why
chunk-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 10d 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 — 308 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Chunk PR
Break one big PR into a sequence of small, logical, merge-safe PRs. Suggest, don't decide — the user approves every chunk and every destructive action.
Preferences
Read ~/.claude/skills/chunk-pr/preferences.md using the Read tool. If not found, no preferences are set.
Context
On startup, use Bash to detect: current branch (git branch --show-current), base upstream, commits ahead, total lines changed (git diff {base}...HEAD --shortstat), and repo (gh repo view --json nameWithOwner -q .nameWithOwner). Skip any that fail. Do NOT run destructive commands at this stage.
Command routing
Check $ARGUMENTS:
help→ display help then stopconfig→ interactive setup then stopreset→ delete~/.claude/skills/chunk-pr/preferences.md, confirm, stop- anything else (including empty) → run the chunking workflow
Flags (parsed from $ARGUMENTS)
--base <branch>— override target base (default: auto-detect or preference)--max-lines <N>— target max lines per chunk (default: 400 or preference)--strategy <kind>—conservative(more/smaller),balanced(default),aggressive(fewer/larger)--dry-run/--plan-only— stop after presenting the plan; don't create branches or PRs--draft— open chunks as draft PRs (default)--ready— open chunks as ready-for-review (overrides--draft)--no-push— create local branches only, skip push/PR creation--no-linear— skip Linear linking- Positional — PR number (e.g.
234), PR URL, branch name, or commit range (main..HEAD). If omitted, use current branch.
Help
Chunk PR — Split a big PR into a sequence of smaller, merge-safe ones
Usage:
/chunk-pr Analyze current branch
/chunk-pr <pr-number> Analyze a specific PR
/chunk-pr <pr-url> Analyze a PR by URL (same or other repo)
/chunk-pr <branch> Analyze a local branch
/chunk-pr main..HEAD Analyze a commit range
/chunk-pr --max-lines 300 Prefer chunks under 300 lines
/chunk-pr --strategy conservative More, smaller chunks
/chunk-pr --dry-run Plan only — don't touch git
/chunk-pr config Set preferences
/chunk-pr reset Clear preferences
/chunk-pr help This help
What it does:
1. Identify the subject PR / branch / range
2. Classify commits and files into logical groups
3. Propose an ordered, dependency-aware split plan
4. Let you approve, reorder, merge, or re-split chunks
5. On approval, create branches + draft PRs + Linear links
6. Leave the parent branch untouched
Current preferences:
(shown above under Preferences)
What ships with it
1 file 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.
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.
- 10d ago First seen · 308 lines · 0 tokens per session scan A 3d41107f583c
chunk-pr is a skill published in the GitHub repository mostafa-drz/claude-skills (4 stars, last pushed 28d ago), licensed MIT. It adds 4 tokens to every session and 3,967 once invoked, about $0.0000 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.
Other skills, from other repositories
git-pushing
Stage, commit, and push git changes with conventional commit messages. Use when user wants to commit and push changes, mentions pushing to remote, or asks to save and push their work. Also activates when user says "push changes", "commit and push", "push this", "push to github", or similar git workflow requests.
iterate
Diverge-converge workflow: spin up N agents on isolated worktrees, each exploring a different approach to the same problem. Compare results side-by-side (with optional preview ports), pick a winner, merge it back. Use when the user says /iterate, 'try multiple approaches', 'explore different directions', or wants to…
bootstrap-project
Use when setting up a new project with the full opinionated workflow — session memory, commit hooks, copyright headers, ROADMAP, and CLAUDE.md. Invoke with "bootstrap project", "setup project", "new project setup", or /bootstrap-project.
session-end
Use when the user says "end session", "wrap up", "let's stop here", or invokes /session-end.
metrics
Computes development quality and velocity metrics from git history and tracks improvement over time. Measures fix ratios, rework hotspots, test coverage, DORA metrics, and code churn.
recall
Reconstructs the development cycle from git history, distills sequential/parallel patterns, and produces actionable insights for improving future iterations. Triggers: recall, retrospective, development analysis, what happened, dev cycle analysis.