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.
git clone --depth 1 https://github.com/emadd/mise-en-claudeWrote 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/emadd/mise-en-claude/mise-handoff)<a href="https://agentmods.dev/commands/emadd/mise-en-claude/mise-handoff"><img src="https://agentmods.dev/badge/commands/emadd/mise-en-claude/mise-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/commands/emadd/mise-en-claude/mise-handoff"><img src="https://agentmods.dev/badge/commands/emadd/mise-en-claude/mise-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 | $0.00038 | $0.00751 |
| Opus 5.5 | $0.00015 | $0.00300 |
| Sonnet 5.5 | $0.00008 | $0.00150 |
| Haiku 4.5 | $0.00004 | $0.00075 |
Grade A, and why
mise-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.
What it actually says
Context is a budget, and this is the hand-off that beats silent auto-compaction. Capture the current state so a fresh session resumes with no loss, then stop, rather than pushing into compaction where fidelity quietly dies.
Write these six fields, tight and specific (focus: $ARGUMENTS if given):
- Goal — what we're actually trying to accomplish (the outcome, not the next keystroke).
- Done — what's finished and verified, with the evidence (a passing test, a running screen). Re-verify it now, live — rerun the test, reload the screen — don't transcribe an earlier claim (yours, a notes file's, a teammate's) on trust. If you can't re-verify something before the hand-off, write it as unverified rather than done; a stale "done" is worse than an honest gap, because the next session builds on top of it without rechecking.
- Next — the very next concrete step, then the ones after it.
- Key decisions — choices made and why, so they aren't relitigated or accidentally undone.
- Files touched — the files that matter and what changed in each; flag anything half-done.
- Gotchas — traps, dead ends, environment quirks, "don't do X, it breaks Y."
Pick the durable target, most-programmatic first:
- Check
.mise/state.jsonfor atrackerthe project was set up with. If it names a platform other than GitHub Issues (Jira, Linear, Asana, Trello, etc.) and it's reachable this session (MCP/API/CLI), write the hand-off there — a task or comment taggedhandoff— so it lives where the rest of the project's work already does. - Otherwise, if
ghis authenticated and the repo has a remote, open a GitHub issue (gh issue create) — it's shared, linkable, and survives the machine. Label ithandoff. - Otherwise write
HANDOFF.mdat the repo root. If you commit it, honor any no-auto-commit rule: ask first, then run the commit yourself on a yes — don't hand over a command to type. - Never leave the hand-off only in chat; chat is the thing about to be lost.
Then give the user a ready-to-paste resume prompt, not just a pointer to the artifact. Put it
in its own fenced code block, on one line, so it can be copied with a single click and dropped
straight into a fresh session as the next /mise-cook (or plain) prompt — e.g.:
Resume from issue #190: <one-line goal>. Read the issue for full context, then continue from "Next".
Don't make the reader assemble this themselves from prose — the whole point of the hand-off is that a fresh session (or a fresh reader) can go from "here's the artifact" to "working again" without hand-crafting anything. Keep the hand-off itself scannable — a fresh reader should grasp the state in under a minute, not wade through a transcript.
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 First seen · 47 lines · 38 tokens per session scan A 48128830984b
mise-handoff is a command published in the GitHub repository emadd/mise-en-claude (51 stars, last pushed 3mo ago), licensed MIT. It adds 38 tokens to every session and 751 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
merge-conflict-analysis
You are analyzing merge conflicts for PR #${{ pr-number }}.
release-notes
Generate consistent, well-structured release notes from git history. Triggered on release tags following semver patterns (v..) to produce categorized changelog with breaking changes, features, fixes, and contributor attribution.
commit
A command that examines staged Git changes and proposes a commit message using the Conventional Commits format, such as feat, fix, or docs. Git is a tool for tracking changes to code.
repo-audit
Audit a codebase (local or remote GitHub/GitLab) against architecture principles and requirements, surfacing drift, risk, and missing decisions.
release-plan
A release-planning command for a software change. It documents the release scope, rollback plan, gradual or full rollout, and monitoring before waiting for human approval.
codex-review
Use when asked to check, turn on, turn off, or switch this repo's Codex review mode (local, github, or off), or to report the effective mode and Codex sandbox status.