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 agentmods add commands/cwakim/claude-plugins/mergegit clone --depth 1 https://github.com/cwakim/claude-pluginsWrote 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/commands/cwakim/claude-plugins/merge)<a href="https://agentmods.dev/commands/cwakim/claude-plugins/merge"><img src="https://agentmods.dev/badge/commands/cwakim/claude-plugins/merge.svg" alt="Measured on agentmods" 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 | $0.00064 | $0.02611 |
| Opus 5 | $0.00032 | $0.01306 |
| Sonnet 5 | $0.00013 | $0.00522 |
| Haiku 4.5 | $0.00006 | $0.00261 |
Grade A, and why
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 3d 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 — 182 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Merge (merge [--dry-run] [path])
Converges this machine's live state with the other machines'
mirrors. Merge reads one, several, or all of the other
machines/<host>/ subtrees from a backup source and applies their content
into the live stores here, with restore's temperament throughout: plan
first, conflicts asked, nothing local ever deleted, nothing applied before
the plan is confirmed. It never writes into any machines/ subtree; the
every-machine-writes-only-its-own invariant from docs/layout.md stands.
It works for any number of machines. Full convergence is one cycle: each
machine in turn merges from all the others and backs up, in any order;
after the last machine's turn every subtree agrees, and resolutions made
early in the cycle propagate through the mirror, so later machines mostly
just confirm them.
Merge covers memories, handoffs, and plans. Config is excluded by
design: settings.json, keybindings, commands, and agents legitimately
differ per machine, and merging them invites subtle breakage. Config moves
between machines only through an explicit restore, where the user has
deliberately chosen a source machine to copy from.
The mirror tree and name-to-target derivation rules live in
${CLAUDE_PLUGIN_ROOT}/docs/layout.md; read it before building the plan.
If the source machine picked below turns out to be this machine's own
subtree, refuse and point at restore: merging a mirror into the machine
it came from is what restore is for.
Resolve the source
Identical to restore's source resolution (see commands/restore.md), except
merge wants every machine's subtree, not just this host's:
<path>given: a zip made byzip <path>, or its already-extracted folder. A.zipfile is extracted into a fresh temp directory (mktemp -d); a directory is used directly. Verify the source root contains at least onemachines/<hostname>/subtree; if not, say so and stop. No GitHub, nogh, no network.- No
<path>: a configured backup target; both kinds may exist.- GitHub mirror (
~/.claude/memory-backup/clone):git pull --ff-onlyfirst so it is current. The source root is the clone. - Object storage (
~/.claude/memory-backup-obstore.json): materialize it into a fresh temp directory with${CLAUDE_PLUGIN_ROOT}/scripts/obstore-pull.sh --dest <tmp> --bucket ...reading bucket/prefix/endpoint/region/profile from the config. Pass no--host, so the wholemachines/tree (every machine) comes down, which is what merge reads across. The temp dir is then the source root, handled exactly like an extracted zip. - Both configured: ask via
AskUserQuestionwhich to merge from. Only one: use it without asking. If nothing is configured, say so and stop: merge presumes an established mirror (for disaster recovery onto a fresh machine, userestore).
- GitHub mirror (
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.
- 3d ago First seen · 182 lines · 64 tokens per session scan A 46018af7ff4d
merge is a command published in the GitHub repository cwakim/claude-plugins (2 stars, last pushed 13d ago), licensed MIT. It adds 64 tokens to every session and 2,611 once invoked, about $0.0003 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 commands, from other repositories
minutes-ideas
Surface recent voice memos and ideas captured from any device. Use when the user asks "what ideas did I have?", "what were my recent memos?", "what did I record while walking?", or wants to recall a captured thought.
learn
Force claude-smart to extract learnings from this session now.
session-end
I'll summarize this coding session and update the memory system with our accomplishments.
memory-store
Store an insight, decision, or pattern to memory.
cc-memory
Configure persistent memory that survives across sessions using a layered approach: split rule files for always-loaded context, auto-memory for organic learning, and optional MCP-backed long-term memory for large codebases.
neo-review
Get Neo's code review with semantic matching against past findings in memory. Use on a diff or module before merge, especially where earlier mistakes in this codebase are likely to recur. Skip for formatting, lint-catchable issues, and single-line changes.