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 skills add mohitbadwal/ringback --skill watchdoggit clone --depth 1 https://github.com/mohitbadwal/ringbackWrote 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/mohitbadwal/ringback/watchdog)<a href="https://agentmods.dev/skills/mohitbadwal/ringback/watchdog"><img src="https://agentmods.dev/badge/skills/mohitbadwal/ringback/watchdog/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/skills/mohitbadwal/ringback/watchdog"><img src="https://agentmods.dev/badge/skills/mohitbadwal/ringback/watchdog.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 | $0.00096 | $0.04165 |
| Opus 5.5 | $0.00038 | $0.01666 |
| Sonnet 5.5 | $0.00019 | $0.00833 |
| Haiku 4.5 | $0.00010 | $0.00417 |
Grade A, and why
watchdog 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 1mo 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 — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Watchdog — autonomous, presence-aware escalation
Turns this session into a watchdog: each tick it checks a watched condition, and when that condition needs you, it decides whether and how to reach you based on whether you're at the laptop. It never interrupts you while you're actively typing; it escalates only after you've gone idle and stayed idle through a short confirmation window.
Escalation ladder (how far it climbs is capped by the declared priority):
chat status → chat WARNING → phone ALERT (one-way) → phone CALL (two-way)
always user away away + warned away + still silent
priority=low→ chat only, never alert/call.priority=medium→ may climb to phone alert (one-way SIP message / push). Default.priority=critical→ may climb to phone call (live conversation).
Backends: mcp__ringback-alert__alert_me (one-way) and mcp__ringback-voice__call_start + converse (two-way) — the two MCP servers from this repo.
State file
All cross-tick state lives in ~/.claude/watchdog/state.json. Read it at the start of every tick; write it before finishing. Shape:
{
"task": "human description of what is being watched",
"condition": "exactly how to check it each tick (command/API + what counts as 'needs me')",
"priority": "medium",
"armed_at": "2026-05-24T10:00:00Z",
"rung": "monitoring", // monitoring | warned | alerted | called | resolved
"warned_at": null, // ISO time the chat WARNING was posted
"last_idle_seconds": 0,
"notes": "" // anything to carry forward (run IDs, etc.)
}
rung meanings: monitoring = nothing fired yet; warned = warning posted, in the 1–2 min confirmation window; alerted = one-way alert sent; called = phone call placed; resolved = watch is fully done and the user has been informed — stop looping. Note: a call where the user hands over a NEW task does NOT go to resolved — it rewrites task/condition and goes back to monitoring (see "Resolving vs. pivoting after contact").
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.
- 1mo ago First seen · 210 lines · 96 tokens per session scan A 034da1f18cce
watchdog is a skill published in the GitHub repository mohitbadwal/ringback (25 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 96 tokens to every session and 4,165 once invoked, about $0.0004 per session on Opus 5.5. 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 skills, from other repositories
google-workspace
Gmail, Calendar, Drive, Docs, Sheets via gws CLI or Python.
simplify-code
Parallel 4-agent cleanup of recent code changes.
first-task
Run the first task chat that setup hands off.
initiate-setup
Run the first-run setup chat in the Hermes desktop app.
himalaya
Himalaya CLI: IMAP/SMTP email from terminal.
apple-reminders
Apple Reminders via remindctl: add, list, complete.