handover

A session handover command that records important context and unfinished work for the next coding session.

In plain words
What is it for?
It helps sort session topics into follow-up actions, project documentation, memory, or a handover file.
Why use it?
It prevents decisions, pending tasks, and useful context from being lost when a session ends.

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/florini/florians-claude-code-kit/handover
Clone the repo
git clone --depth 1 https://github.com/FlorinI/Florians-Claude-Code-Kit
Per session 61 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,171 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 $0.00061 $0.03171
Opus 5 $0.00030 $0.01586
Sonnet 5 $0.00012 $0.00634
Haiku 4.5 $0.00006 $0.00317

Measured 2d ago against content hash 1eeddb7f4182, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

handover 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 2d 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.

home/commands/handover.md · 184 lines

How it starts

The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/handover — deliberate session handover

You are about to do a clean session handover. Don't /clear yet; the user will do that after the handover file is written.

Invocation — /handover vs /handover ok

If the command was invoked with the argument ok (case-insensitive — ok, OK, Ok all count), run the flow with the proposed default buckets and no confirmation prompt — but NOT silently. Specifically, in this order:

  1. Draft the eight coordinates (Step 1) internally. You do not need to print the long-form draft.
  2. Print the Step-2 triage summary table — the compact N. [bucket] item list, defaults already filled in, in the exact format the interactive flow uses (see Step 2). Show it BEFORE acting on or writing anything. This is the recap the user reacts to: because it lands before any M/D/X action or file write, the user can scan it and hit Esc to interrupt and correct course if a bucket is wrong.
  3. Do NOT pause for triage adjustments and do NOT ask for confirmation. After printing the table, continue straight on in the same turn — act on the M, D, and X items (Step 3), write the handover file (Step 4), and report (Step 5).

The only pauses still permitted are the deliberate-action confirmations in Step 3 for outward-facing or hard-to-reverse X items (commits, pushes, deletes, sends) — those are never auto-run off defaults.

Otherwise (plain /handover), run the full interactive flow below: present the draft, take the user's triage adjustments, then act.

Step 1 — draft the eight coordinates

Without reading any new files, draft the handover by filling in the sections below from your conversational context. Be concrete and specific — "Discussed the rewrite" is weak; "Agreed lesson 5 opens at the gateway-ASIN question; AMC source still unsourced" is right.

  1. Goal — What are we trying to accomplish? One or two sentences naming the objective and why it matters.
  2. What was tried — What approaches were explored this session, especially any that failed or were abandoned and why.
  3. Decisions & constraints discovered — Choices that were made ("X over Y because Z") and constraints that surfaced ("the API doesn't support batch mode") that aren't captured in code or git history.
  4. Recent changes — What was actually modified this session. Concrete: file paths, what changed in each, commit hashes if any.
  5. Files in flight — Files still being actively edited or that have uncommitted changes.
  6. Next steps — What to do next, in priority order. Note blockers inline ("step 3 is blocked on user confirming the schema").
  7. User preferences — Working-style signals that surfaced this session: communication preferences, conventions the user corrected you on, patterns they validated. Only things the next session's Claude would otherwise have to re-learn.
  8. Other context — Anything that doesn't fit the sections above but would be useful to the next session: half-formed hypotheses, anomalies noticed but not pursued, tangential observations.

Read the full file on GitHub · 184 lines

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. 2d ago First seen · 184 lines · 61 tokens per session scan A 1eeddb7f4182

Subscribe to this mod's changes

handover is a command published in the GitHub repository FlorinI/Florians-Claude-Code-Kit (5 stars, last pushed 4d ago), licensed MIT. It adds 61 tokens to every session and 3,171 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.