Borrowing it
Nothing to install: this file belongs to hlop3z/claude-ai-dangerous. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/hlop3z/claude-ai-dangerous/main/.claude/commands/ai/handoff.mdgit clone --depth 1 https://github.com/hlop3z/claude-ai-dangerousWrote 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/hlop3z/claude-ai-dangerous/handoff)<a href="https://agentmods.dev/commands/hlop3z/claude-ai-dangerous/handoff"><img src="https://agentmods.dev/badge/commands/hlop3z/claude-ai-dangerous/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/hlop3z/claude-ai-dangerous/handoff"><img src="https://agentmods.dev/badge/commands/hlop3z/claude-ai-dangerous/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 | $0.00018 | $0.00717 |
| Opus 5 | $0.00009 | $0.00358 |
| Sonnet 5 | $0.00004 | $0.00143 |
| Haiku 4.5 | $0.00002 | $0.00072 |
Grade A, and why
AI: 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 10d 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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Generate a HANDOFF.md at the project root that snapshots the current session so a future
session can resume without context loss.
This is a snapshot, not a log. Each run produces the current state of the world — it is a resume note for picking your own work back up, not a historical record.
Hard rules
- Overwrite, never append. If
HANDOFF.mdalready exists, replace it entirely. Never accumulate prior sessions' content — an append-mode file rots and contradicts itself. - Reference, never duplicate. For anything that already has a canonical home, write a
pointer to that home — do not copy its contents. This preserves single source of truth
and keeps the snapshot from drifting:
- Active changes →
openspec list(name them, link toopenspec/changes/<name>/) - Open work → the relevant
tasks.md - Decisions + rationale → the relevant
design.mdADR section - Durable facts about the user/project → the memory dir
- Active changes →
- Only capture what has no other home. Free-form prose is restricted to the three things below. Everything else is a reference.
- Gitignored. Ensure
HANDOFF.mdis in.gitignore; add it if missing. The snapshot is a private resume note, not a committed canonical home. - Promote, don't hoard. If the session produced a durable finding (a decision, a spec,
a fact), promote it to its proper home — don't let
HANDOFF.mdbecome its storage.
Steps
- Run
openspec list --json(if the project uses OpenSpec) to enumerate active changes to reference. Skip silently if not an OpenSpec project. - Check
.gitignoreforHANDOFF.md; add the entry if absent. - Write
HANDOFF.mdat the project root using the structure below, OVERWRITING any existing file. - Report the path written and confirm it is gitignored.
HANDOFF.md structure
# Handoff
> Session snapshot — overwritten each run. Resume state only; canonical content lives in
> the linked sources, not here.
## Session narrative
<!-- What this session was about and what got done, across changes. The story that no
single tasks.md or design.md tells. Keep it tight. -->
## Current state / where work stopped
<!-- In-flight state: what is half-done, what is blocked, the exact point work paused. -->
## Next steps
<!-- The recommended next actions for the future session, in priority order. -->
## References
<!-- Pointers ONLY — no copied content.
- Active changes: openspec/changes/<name>/
- Open work: <path>/tasks.md
- Decisions: <path>/design.md
- Durable facts: memory/ -->
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.
- 10d ago First seen · 76 lines · 18 tokens per session scan A e7f789d3d80f
AI: Handoff is a command published in the GitHub repository hlop3z/claude-ai-dangerous (2 stars, last pushed 25d ago), licensed MIT. It adds 18 tokens to every session and 717 once invoked, about $0.0001 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
load
Session lifecycle management with Serena MCP integration for project context loading.
save
Session lifecycle management with Serena MCP integration for session context persistence.
INFO
A startup routine for giving a coding agent the context it needs at the beginning of a terminal session. It checks platform, project, and current-task information from a fixed set of files and records a short summary.
mpm-session-resume
Load context from paused session.
resume
Resume work from .claude/handoff/current.md in this fresh session (skip --resume).
save
Save a compact handoff checkpoint so a fresh session can resume cheaply (no --resume).