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 lacerbi/dotclaude --skill handoffgit clone --depth 1 https://github.com/lacerbi/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/lacerbi/dotclaude/handoff)<a href="https://agentmods.dev/skills/lacerbi/dotclaude/handoff"><img src="https://agentmods.dev/badge/skills/lacerbi/dotclaude/handoff/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/lacerbi/dotclaude/handoff"><img src="https://agentmods.dev/badge/skills/lacerbi/dotclaude/handoff.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 | $0.00047 | $0.01059 |
| Opus 5.5 | $0.00019 | $0.00424 |
| Sonnet 5.5 | $0.00009 | $0.00212 |
| Haiku 4.5 | $0.00005 | $0.00106 |
Grade A, and why
handoff 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 1mo 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/handoff — make the work resumable from a clean checkout
Goal: a fresh git clone plus a read of the project's durable docs is enough for the next session to
resume — except artifacts a project keeps local by design (gitignored checkpoints, pools, weights,
venvs; .gitignore is the source of truth). Nothing meant to be tracked may live only in this
conversation, the working tree, or a session-bound process. Close every gap you find — don't just report it.
1. Capture the session's substance (if any emerged)
Decisions and their rationale, findings/results, approaches tried and rejected, open questions — anything from the discussion worth keeping that isn't already in the code — into the project's durable docs (DEVLOG / TODO / notes, whatever the project uses). The conversation is not durable: a decision or finding that lives only in chat is lost. Skip if nothing substantive emerged or if it has already been recorded.
2. Record in-flight / async work
For running or pending work whose continuation depends on session-specific context — cluster jobs, background processes, scheduled tasks, or remote work not otherwise discoverable — record in an active durable doc its ID, the exact command/config, where outputs land, how to retrieve them, and its status, so a reader can resume it without this conversation.
An open PR is already durable and discoverable remote state. Do not copy its number, URL, transient status, or CI state into a plan, devlog, TODO, or completed/archive record merely because it is open; report those details in the final reply. Add repository documentation only when the project already maintains a PR tracker or when resumption requires non-obvious information that the PR and repository do not contain. Keep such information in an active work record, not a completed or archived one.
3. Surface what is NOT durable
- remote/scratch artifacts not yet collected; local-only artifacts the next session must regenerate or fetch (say how);
- session-bound watchers / background tasks that die with this session (the underlying job keeps running — say how to re-attach by ID);
- auth / secrets / local state that won't carry over.
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.
- 1mo ago Changed · +6 lines 137604275333
- 1mo ago First seen · 73 lines · 47 tokens per session scan A 88b01a0310cf
handoff is a skill published in the GitHub repository lacerbi/dotclaude (2 stars, last pushed yesterday), licensed MIT. It adds 47 tokens to every session and 1,059 once invoked, about $0.0002 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
archaeologist
Use when changing unusual or defensive code, mature APIs, dependencies, or architecture, or estimating a migration where Git history may reveal hidden constraints.
committing-code
Writes git commit messages using conventional commits format with gitmoji. Use when creating git commits, preparing commit messages, or when the user asks to commit changes. Triggers on "commit", "git commit", "save changes", or any request to record changes to version control.
addressing-pr-feedback
Fetches, organizes, and addresses PR review comments from GitHub. Use when user asks to review PR comments, fix PR feedback, check what reviewers said, address review comments, or handle bot suggestions on a pull request. Triggers on "review PR", "fix comments", "PR feedback", "what did reviewers say", "address PR…
datalad
Retrieves, versions, and publishes scientific datasets with DataLad and git-annex, and captures computational provenance with datalad run, rerun, and containers-run. Use when cloning or fetching data from OpenNeuro, DANDI, datasets.datalad.org, or any DataLad dataset; when a file in a dataset reads as a broken symlink…
git-ai-search
Search and restore AI conversation context from git history.
mcore-split-pr
Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.