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/beevibe-ai/beevibe-ctoWrote 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/beevibe-ai/beevibe-cto/guard)<a href="https://agentmods.dev/commands/beevibe-ai/beevibe-cto/guard"><img src="https://agentmods.dev/badge/commands/beevibe-ai/beevibe-cto/guard.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 5.1 | $0.00050 | $0.00577 |
| Opus 5 | $0.00025 | $0.00289 |
| Sonnet 5 | $0.00010 | $0.00115 |
| Haiku 4.5 | $0.00005 | $0.00058 |
Grade A, and why
guard 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 8d 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 — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/adr:guard — Install write-time + commit-time hooks
Use when the user wants the bot to enforce their team's principles automatically, not just on demand.
Step 0 — Check principles exist
If .adr/principles.json doesn't exist:
Your team principles haven't been discovered yet. Run
/adr:principlesfirst — the hooks need somewhere to read the rules from.
Stop.
Step 1 — Install the hooks
adr guard install
This writes two files:
.claude/settings.local.json— adds a PreToolUse hook on Edit/Write/MultiEdit. Every time Claude Code is about to write to a file, the hook checks the team principles for that file's path and injects the relevant ones into the agent's context. No LLM call — pure file-based lookup..git/hooks/pre-commit— runsadr review --staged --top-n 5before each commit. Blocks the commit if any HIGH-severity violation is found; lets medium/low through as advisory.
Both installs are idempotent — running again is a no-op.
Step 2 — Show the user what's now in effect
The hooks are installed. From now on:
- At write time, every Edit/Write tool call shows me the team principles for that file's area. I'll surface conflicts if the change would violate one.
- At commit time,
git commitruns an automatic check. HIGH-severity violations block the commit. Bypass withgit commit --no-verifyif you must.
Step 3 — Offer the dry-run
Want to test it? Tell me to edit a file in this repo and I'll show you the principles that fire.
Notes for the bot
- The pre-write hook is silent when no principles match. That's intentional — write-time noise on every Edit kills the signal. Only fires when a discovered example shares a top-level dir with the edited file, or the lens is broadly applicable.
- Re-run
adr guard installto refresh. Principles can change (re-running/adr:principlesupdates them); the install command is idempotent so re-running is safe. - Pre-commit failures are recoverable. If the user hits a block they disagree with, they can either (a) fix the issue, (b) update the principle via
/adr:principles, or (c) bypass with--no-verifyfor an emergency commit.
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.
- 8d ago First seen · 46 lines · 50 tokens per session scan A c02779a2a005
guard is a command published in the GitHub repository beevibe-ai/beevibe-cto (105 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 50 tokens to every session and 577 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 2026-08-30.
Other commands, from other repositories
diff-script
Compare an MDL script against the project's current state.
commit
A command that examines staged Git changes and proposes a commit message using the Conventional Commits format, such as feat, fix, or docs. Git is a tool for tracking changes to code.
commit
Stage, commit, and optionally ship further. Pass an escalation token (push, pr, merge, squash, squash merge) to skip the prompt. With no token, commits then asks how far to ship. Delegates message format to the atomic-git-discipline skill.
afsr
Review feature spec - improve the spec itself before implementation (shortcut for feature-spec-review).
next
Suggest the most likely next workflow action based on current context.
commit.zh-CN
A command for creating a Git commit, which is a saved checkpoint of changes in a Git project.