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 MithunWijayasiri/dotclaude --skill git-mergegit clone --depth 1 https://github.com/MithunWijayasiri/dotclaudeWrote 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/mithunwijayasiri/dotclaude/git-merge)<a href="https://agentmods.dev/skills/mithunwijayasiri/dotclaude/git-merge"><img src="https://agentmods.dev/badge/skills/mithunwijayasiri/dotclaude/git-merge/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/mithunwijayasiri/dotclaude/git-merge"><img src="https://agentmods.dev/badge/skills/mithunwijayasiri/dotclaude/git-merge.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 | $0.00081 | $0.02482 |
| Opus 5.5 | $0.00032 | $0.00993 |
| Sonnet 5.5 | $0.00016 | $0.00496 |
| Haiku 4.5 | $0.00008 | $0.00248 |
Grade A, and why
git-merge 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 18d 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 — 118 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Git Merge — Safe Branch Integration
Workflow the main agent runs. Investigate before mutating; preserve local-only unstaged changes; resolve conflicts by diagnosis; verify with the repo's own checks before done.
Scripts in this skill's scripts/ folder (bash, run with cwd = repo root, as bash <skill-dir>/scripts/<name>.sh) do the mechanical steps and print compact digests. Below, scripts/ means that folder.
Hard rules
- Read-only investigation first.
merge-recon.sh(fetch + map state) before anystash/merge. - Never auto-commit beyond the one merge the user authorized.
git commit/--amendneed explicit user approval — ask (amend / separate fixup / leave it). git add/stash/mergeare mutating — explain, then run (user already asked to merge = authorization for the merge itself).- Keep the user's local-only unstaged files OUT of any commit. Stage only merge-resolution files.
- Never push unless asked.
- Never install tooling mid-merge. Run only checks the repo already has.
1. Recon
bash scripts/merge-recon.sh <target> # default target: origin's default branch; current branch from HEAD
Runs git fetch origin (all of origin, not one branch: fetch origin <target> leaves origin/<current-branch> outdated, and recon compares both), then prints one digest: target commits + merge-base, ahead/behind OWN remote (+ partial remote merge), files both merges bring, local unstaged/untracked/ignored, COLLISIONS. Dies on merge/rebase in progress, unmerged paths, detached HEAD, missing origin/<target>.
Why:
- Local branch may be behind its own remote, remote may already hold a partial
Merged <target>commit → "merge properly" = two steps: sync (--ff-only) toorigin/<current-branch>, then mergeorigin/<target>. Decide before touching tree. - Ignored files matter. Plain
git statushides them;mergeoverwrites an ignored file silently, exit 0 when incoming side tracks that path (local.env/config lost). Script uses--ignored. - Both merges checked.
--ff-onlysync rewrites tree too → same collisions. - COLLISIONS: modified → stash (step 3); untracked → merge aborts, move aside; ignored → silent overwrite, move aside. None → no stash needed.
- Lists are capped (
+N more), counts exact. Merge attempt names every conflict; trust it over lists.
What ships with it
5 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.
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.
- 18d ago Changed b6c63666de21
- 19d ago Changed · +5 tokens per session 02c3605bbc49
- 20d ago Changed · -43 lines 8a4d2e01b033
- 2mo ago First seen · 161 lines · 76 tokens per session scan A 5a7e009da159
git-merge is a skill published in the GitHub repository MithunWijayasiri/dotclaude (4 stars, last pushed 4d ago), licensed MIT. It adds 81 tokens to every session and 2,482 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
git-worktree
Use this skill whenever a task involves running parallel work on a git repository without conflicts — e.g. "fix this bug while I keep working on my feature", "run Claude on two branches at once", "work on this task in isolation", "create a throwaway branch for this experiment", or anything where two agents or sessions…
okhp3-repository-organizer
OverKill Hill P³ repository organizer for content-first Git repositories. Use when a local Git repository grew organically from a folder and needs an evidence-based profile, purpose summary, classification, cross-platform naming cleanup, folder design, governance scaffolding, or a safe reorganization plan. Also…
okhp3-repl-repo-janitor
Clean up a single Replit-hosted Git repository — squash-merge and delete old, merged, or abandoned branches and pull requests (including Replit-generated subrepl-, replit-agent, and agent/ branches), and normalize file/folder names against a kebab-case default with documented exceptions. Use when a user asks to purge…
okhp3-github-skill-foundry
Design, author, evaluate, and install-plan a task-focused GitHub Copilot Agent Skill. Use when creating a repository or personal GitHub Copilot SKILL.md for coding, review, maintenance, or repository workflows. Do not use for Cowork, SharePoint, Copilot Studio, or general agent instructions.
okhp3-replit-build-in-public
Full playbook for a Replit Designathon competition submission — canvas organisation for judges, artifact selection, dual deployment (Replit Publish + GitHub Pages), session-to-session narrative continuity, and building-in-public commit discipline. Activate for "designathon", "competition", "building in public"…
okhp3-replit-repl-janitor
OverKill Hill P³ one-Repl repository cleanup workflow for safely auditing and tidying a single Replit workspace's Git checkout. Use when a user asks to clean up merged or abandoned branches and pull requests, purge stale subrepl- or agent/ branches, normalize file and folder names, or remove repository detritus. Also…