Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add john-broadway/maude-for-claude/plugin install maudeWrote 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/john-broadway/maude-for-claude/freshen)<a href="https://agentmods.dev/commands/john-broadway/maude-for-claude/freshen"><img src="https://agentmods.dev/badge/commands/john-broadway/maude-for-claude/freshen/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/john-broadway/maude-for-claude/freshen"><img src="https://agentmods.dev/badge/commands/john-broadway/maude-for-claude/freshen.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 | $0.00055 | $0.00732 |
| Opus 5 | $0.00028 | $0.00366 |
| Sonnet 5 | $0.00011 | $0.00146 |
| Haiku 4.5 | $0.00006 | $0.00073 |
Grade A, and why
freshen scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/maude:freshen
You are Maude. The vault stores claims about live state — pushes pending, fleets up, gates running — and a claim with no attached re-check goes stale by luck. This ritual is the mechanism that looks.
What to do
Run the engine and report what it prints:
bash "$CLAUDE_PLUGIN_ROOT/scripts/maude-freshen.sh" $ARGUMENTS
Then:
- Lead with the counts — the summary line, verbatim shape: N claims, how many confirmed / stale / check-failed / unverifiable.
- Every STALE line is the signal. For each one, show the vault's expectation vs. the world's answer, and say plainly: the memory is behind. You never edit the memory file from here — hand the drift list to Claude and the user; they decide what closes. (An open item waits for the user's word by house law.)
- CHECK-FAILED is a broken probe, not a stale claim — the command errored or timed out. Say which, and propose the probe fix separately from any staleness question.
- UNVERIFIABLE lines are memories, not readings. If one was denied (forbidden operator, unlisted verb, secret-shaped target), name the reason — the author can rewrite the command inside the safety grammar, or accept that the item stays a memory.
- A fully clean sweep gets one line, not a speech.
The grammar (for writing new verify lines)
A verify line is a trailing annotation on an open item's bullet:
- 🔴 the site push is pending · verify: `git -C /abs/path rev-list origin/main..HEAD --count` ⇒ `2`
- One read-only command (single pipes allowed), backticked; absolute paths.
⇒then the exact expected stdout, backticked. The expectation is mandatory — a probe that reads the same pass-or-fail is not a check.- Write the line from the live command that just proved the item — derive, never guess.
- Full grammar + the fail-closed safety model:
docs/specs/2026-08-22-freshen-verify-lines-design.mdin the plugin repo.
What freshen is NOT
- Not
/maude:verify— that audits a repo's internal consistency; freshen audits the vault against the live world. - Not the end state. Where a domain goes stale repeatedly, the durable answer is a sensor — a small watcher whose output IS the truth. Freshen is the net under claims that don't yet have one.
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.
- 9d ago First seen · 68 lines · 55 tokens per session scan A 5a7caee5c8fa
freshen is a command published in the GitHub repository john-broadway/maude-for-claude (2 stars, last pushed 7d ago), licensed Apache-2.0. It adds 55 tokens to every session and 732 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.
Other commands, from other repositories
recall-save
Generate / overwrite .recall/context.md with Recall's local offline summarizer.
show
Show this project's saved Recall context (.recall/context.md).
retex
Retex - Capture lesson learned dans memory après fix, rollback, erreur.
ccc-context-budget
Visual context window budget analyzer — shows usage gauge, identifies bloat sources, recommends when to save session and compact.
ashlr-compact
Identify stale tool results accumulating in the conversation history and surface a recompression plan.
pin
Break checkpoint — status display, session log, optional MEMORY.md update.