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/Fredje777/claude-zenvibeWrote 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/fredje777/claude-zenvibe/zenpause)<a href="https://agentmods.dev/commands/fredje777/claude-zenvibe/zenpause"><img src="https://agentmods.dev/badge/commands/fredje777/claude-zenvibe/zenpause/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/fredje777/claude-zenvibe/zenpause"><img src="https://agentmods.dev/badge/commands/fredje777/claude-zenvibe/zenpause.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 | $0.00066 | $0.01023 |
| Opus 5 | $0.00033 | $0.00511 |
| Sonnet 5 | $0.00013 | $0.00205 |
| Haiku 4.5 | $0.00007 | $0.00102 |
Grade A, and why
zenpause 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 — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You produce a complete handoff so the user (or yourself) can resume cleanly hours later.
Output language: Default to English. If CLAUDE.md exists at the project root and is written in French, OR if the existing journal (docs/JOURNAL.md or JOURNAL.md) contains French entries, write your output (confirmation message, new journal entry) in French. Otherwise English. Detect once and stay consistent across the whole skill execution.
Workflow
0. Verify the working directory is a git repo
Run git rev-parse --is-inside-work-tree. If it fails, the project is not under git: skip step 1 and the "Git" section of step 3, write only the journal entry, and mention in the confirmation that no git checkpoint was made.
1. Commit + push everything committable
- Run
git statusandgit diff --statto see what changed. - Run
git log -10 --onelineto learn the project's commit message convention. - For each modified file: if it is in a clean, working state, include it in the commit. If it is clearly WIP or half-written (broken syntax, stub functions, partial refactor), do not commit it — flag it in step 3 instead.
- Stage the clean files and commit with a message that summarizes what was actually done this session, not "WIP".
- If a remote is configured, run
git push. If push fails for a missing upstream, set it (git push -u origin <branch>) and retry. Never use--forceor--no-verify.
2. Locate (or create) the journal
Look for the journal in this order:
docs/JOURNAL.mdJOURNAL.mdat the repo root- If neither exists, create
docs/JOURNAL.md(runmkdir -p docsfirst).
3. Prepend a new entry to the journal
The newest entry goes at the top of the file. Use this structure (translate headings to French if the project is in French):
## YYYY-MM-DD HH:MM — Pause
> Pause note: <text passed as argument, if any>
### Completed tasks (current iteration)
- ...
### Current task
- ... (precise state: what is done, what remains on this task)
### Remaining tasks (in order)
1. ...
2. ...
### Technical decisions made this session
- ...
### Open questions
- ...
### Git
- Branch: <branch>
- Last commit: <short sha> <message>
### Attention points for the resume
- (optional — refactors to do, uncommitted WIP files, tests to write, code smells)
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 · 94 lines · 66 tokens per session scan A 919ef3ce85fc
zenpause is a command published in the GitHub repository Fredje777/claude-zenvibe (2 stars, last pushed 16d ago), licensed MIT. It adds 66 tokens to every session and 1,023 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
status
Show the Salesforce project banner — connected org, edition, API, project metadata stats, and MCP server status.
setup
Verify ClauDex prerequisites — Codex CLI, authentication, and git.
red-team
Stress-test a plan, strategy, PRD, or launch with a room of hostile expert personas before you commit.
acos-checkpoint
Create checkpoints before risky operations. Rollback if things go wrong.
instinct-import
Import a portable YAML instinct pack with confidence-gated merge (Instinct Engine).
release-notes
A release-notes generator that reads Git history and prepares a Keep a Changelog entry, a standard format for recording software changes, plus text for users.