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 skills/anatolykoptev/dozor/dev-modenpx skills add anatolykoptev/dozor --skill dev-modegit clone --depth 1 https://github.com/anatolykoptev/dozorWrote 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/skills/anatolykoptev/dozor/dev-mode)<a href="https://agentmods.dev/skills/anatolykoptev/dozor/dev-mode"><img src="https://agentmods.dev/badge/skills/anatolykoptev/dozor/dev-mode.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 | $0.00077 | $0.01283 |
| Opus 5 | $0.00039 | $0.00642 |
| Sonnet 5 | $0.00015 | $0.00257 |
| Haiku 4.5 | $0.00008 | $0.00128 |
Grade A, and why
dev-mode 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 6d 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 — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dev Mode
Prevent auto-fixing services under active development.
When to Activate
Automatic signals (suggest to user or self-activate)
| Signal | Confidence | Action |
|---|---|---|
| Same service restarted 3+ times in 10 min | High | Exclude that service |
| Triage errors: "connection refused", "no such image", "build" | High | Suggest dev mode for that service |
| Deploy just completed (< 5 min ago) | Medium | Suggest excluding deployed services |
| Multiple services unhealthy simultaneously after being healthy | Medium | Suggest global dev mode |
| User says "I'm working on X" / "rebuilding X" | Certain | Exclude X immediately |
Manual activation via MCP tool
server_dev_mode(enable: true) — global observe-only
server_dev_mode(exclude: ["my-service"], ttl: "2h") — skip a service from triage
server_dev_mode(exclude: ["api-service", "worker"]) — skip multiple (default 4h TTL)
server_dev_mode(status: true) — check current state
Dev Mode Behaviors
Global dev mode ON
- Watch still runs triage on schedule
- Triage report includes
DEV MODE ACTIVEbanner - Watch prompt changes to: "observe only, do NOT take any corrective action"
- Agent MUST NOT restart, deploy, or modify any services
- Agent CAN still diagnose, inspect, and report
Per-service exclusions
- Excluded services are removed from triage entirely
- They still appear in
server_inspect(mode: "health")(unfiltered) - Exclusions auto-expire after TTL (default 4h)
- If all services excluded, triage says "All services dev-excluded"
Decision Flow
Triage detects problems
│
├─ Is dev mode ON?
│ └─ YES → Report only. Do NOT fix. Done.
│
├─ Is the problematic service excluded?
│ └─ YES → Skip it. Not in triage output.
│
├─ Does the error look like active development?
│ │ (connection refused, image missing, build error,
│ │ repeated restarts in short window)
│ └─ YES → Suggest excluding the service:
│ "Service X appears to be under active development
│ (3 restarts in 5 min). Exclude from triage?"
│ If via watch (no user): auto-exclude for 1h,
│ log the decision.
│
└─ Normal incident → follow incident-response skill
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.
- 6d ago First seen · 129 lines · 77 tokens per session scan A dc509c5a471e
dev-mode is a skill published in the GitHub repository anatolykoptev/dozor (5 stars, last pushed yesterday), licensed MIT. It adds 77 tokens to every session and 1,283 once invoked, about $0.0004 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 skills, from other repositories
windows-shell
Manage long-running background jobs and interactive workflows on Windows using PowerShell Jobs (Start-Job, Receive-Job) — the Windows equivalent of tmux sessions.
schedule_manage
Tool: schedulemanage. All ops use action field.
Read, search, and draft email in the connected mailbox (Gmail or Microsoft 365). Use when the user asks about their inbox, wants a message found or summarized, or wants a reply drafted. Sending is gated by a separate, explicitly granted permission — never assume you can send, and never claim a message was sent unless…
library
当要写/整理在线文档、建数据表增删改查、导入 CSV·Excel、做看板/dashboard/运营页/汇报页、md 转网页或演示 HTML 发布、建目录、上传下载网盘文件、审阅修订、分享协作，以及提到资料库/知识库/网盘/空间/workbuddy.cn/space 链接时使用。资料库是 WorkBuddy 原生的内容管理、分享协作与轻发布模块，覆盖在线文档（doc）、网页链接（link）、结构化数据表（database）、数据/汇报页面（page）、网盘文件（drive）与空间管理（manage）；凡说资料库/知识库或出现 space 链接、未显式点名腾讯文档/飞书/Notion/乐享/金山文档/iWiki 等外部产品时一律走本…
automation
Schedule reminders and recurring tasks.
cron
Schedule reminders and recurring tasks.