carry

A handoff command that records what one coding agent changed and where the work is stored, such as its Git worktree and branch.

In plain words
What is it for?
Use it when pausing or finishing work so another agent can continue from the correct repository location and understand the changes.
Why use it?
It removes the need for the next agent to reconstruct the previous agent’s progress from scattered Git details.

Command

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 commands/young1the/pets-driven/carry
Clone the repo
git clone --depth 1 https://github.com/young1the/pets-driven
Per session 40 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 652 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.

ModelPer sessionOnce invoked
Fable 5 $0.00040 $0.00652
Opus 5 $0.00020 $0.00326
Sonnet 5 $0.00008 $0.00130
Haiku 4.5 $0.00004 $0.00065

Measured yesterday against content hash 3df38f0759a6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

carry 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 yesterday.

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.

plugins/pets-driven/commands/carry.md · 62 lines

What it actually says

Hand an agent's work to the next agent. Agents usually work on an isolated branch in a git worktree; when one finishes (or pauses), the next agent needs two things to continue: what was done and where it lives. This skill gathers both and produces one compact, paste-able handoff. Keep it factual and short — the goal is for another agent to read it and immediately know where to go and what state the work is in.

  1. Locate the work. Find the worktree, its source repo, and the base branch:

    git rev-parse --show-toplevel          # this worktree's path
    git rev-parse --git-common-dir         # shared repo (differs in a worktree)
    git worktree list
    git branch --show-current
    

    The base is the branch this work will eventually rejoin (often the branch in the main worktree or the repo's default branch). If it is not obvious, ask rather than guess.

  2. Summarize what was done relative to the base, then turn it into a short, plain-language description — not just raw git output. Read the commit messages and the diff so the summary captures intent, not only filenames:

    git log --oneline <base>..HEAD          # commits added
    git diff --stat <base>...HEAD           # files + line counts
    git status --short                      # uncommitted / untracked work
    

    Flag uncommitted work explicitly — another agent cloning or fetching the branch will not receive changes that are not committed yet.

  3. Write the handoff. Produce a single block in this shape, filled with real values. This is the thing the next agent reads:

    ## Handoff
    Location:    <worktree-path>   (branch <branch>, base <base>)
    Source repo: <source-repo>
    Done:
      - <plain-language summary of what changed and why>
      - <…>
    Commits:     <oneline list, or "none committed yet">
    Files:       <X files, +A/-B>
    Uncommitted: <none | short list>
    Continue:    cd "<worktree-path>"   # already on branch <branch>
    

    Keep "Done" in the agent's own words — that is the part another agent cannot reconstruct from git alone.

  4. Point at integration only briefly. The next agent decides how the work rejoins the base; just name the options so they know the shapes available — fast-forward merge, cherry-pick, or push + PR (all run from the source repo) — without spelling out full command sequences here.

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.

  1. yesterday First seen · 62 lines · 40 tokens per session scan A 3df38f0759a6

Subscribe to this mod's changes

carry is a command published in the GitHub repository young1the/pets-driven (1 stars, last pushed 2d ago), licensed MIT. It adds 40 tokens to every session and 652 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.