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 agentmods add commands/bengous/claude-code-plugins/rewind-summarygit clone --depth 1 https://github.com/bengous/claude-code-pluginsWrote 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/bengous/claude-code-plugins/rewind-summary)<a href="https://agentmods.dev/commands/bengous/claude-code-plugins/rewind-summary"><img src="https://agentmods.dev/badge/commands/bengous/claude-code-plugins/rewind-summary.svg" alt="Measured on agentmods" 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 | $0.00025 | $0.00786 |
| Opus 5 | $0.00013 | $0.00393 |
| Sonnet 5 | $0.00005 | $0.00157 |
| Haiku 4.5 | $0.00003 | $0.00079 |
Grade A, and why
rewind-summary 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.
How it starts
The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Rewind Summary
Produce a compact, copy-ready summary of what was accomplished since a checkpoint, so the user can carry that memory back into the conversation after a native Restore wipes it.
Why this command exists
Native rewind (/rewind, or double-tap Esc on an empty prompt) offers two moves that don't combine:
- Restore repositions to an earlier checkpoint but discards all memory of the turns after it.
- Summarize compresses turns in place but never repositions — it leaves you at the end.
Nothing native does both: go back to an earlier point and keep a summary of what happened since. This command fills that gap.
The wrinkle it must respect: the summary you write is the last message — inside the exact span a Restore throws away. It survives only if the user copies it before rewinding, then pastes it back as the opening of the rewound conversation. So the output is built to be copied first.
You cannot perform the rewind yourself — Restore is a manual user action. Your job is to write the note and hand over the steps.
What to write
The user plans to rewind to this checkpoint: "$ARGUMENTS"
Summarize what was accomplished after it — the memory the rewound conversation needs to continue without redoing or relitigating anything.
If "$ARGUMENTS" is empty, summarize the session's accomplishments so far and remind the user to choose the checkpoint they'll Restore to.
Shape (5–10 lines, hard ceiling)
Capture outcomes, not the journey:
- What was done, fixed, or decided
- Concrete artifacts: files, functions, or bugs touched; commits (relative paths only)
- Findings or constraints that would otherwise be lost
- Test/verification results (pass/fail) if any
- A final line naming the next step — this orients the rewound conversation
Keep it dense. Leave out code snippets, step-by-step narration, reasoning, and absolute paths. If it reads like a transcript or a diff, it's wrong — it's a briefing, not a replay.
How to present it
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.
- yesterday First seen · 73 lines · 25 tokens per session scan A 7db2e1d7038b
rewind-summary is a command published in the GitHub repository bengous/claude-code-plugins (4 stars, last pushed 3d ago), licensed MIT. It adds 25 tokens to every session and 786 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 2026-09-03.
Other commands, from other repositories
OPSX: Onboard
Guided onboarding - walk through a complete OpenSpec workflow cycle with narration.
OPSX: Bulk Archive
Archive multiple completed changes at once.
OPSX: Verify
Verify implementation matches change artifacts before archiving.
OPSX: Continue
Continue working on a change - create the next artifact (Experimental).
OPSX: New
Start a new change using the experimental artifact workflow (OPSX).
workflow
Command "workflow" from n8n-io/n8n, covering workflow, workflow list, workflow get, workflow create and workflow update.