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 Insajin/autopus-adk --skill monitor-patternsgit clone --depth 1 https://github.com/Insajin/autopus-adkWrote 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/insajin/autopus-adk/monitor-patterns)<a href="https://agentmods.dev/skills/insajin/autopus-adk/monitor-patterns"><img src="https://agentmods.dev/badge/skills/insajin/autopus-adk/monitor-patterns.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00016 | $0.01739 |
| Opus 5 | $0.00008 | $0.00870 |
| Sonnet 5 | $0.00003 | $0.00348 |
| Haiku 4.5 | $0.00002 | $0.00174 |
Grade A, and why
monitor-patterns 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 4d 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 — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Monitor Patterns Skill
OMP 2.1.x Monitor tool을 활용한 pane 이벤트 스트리밍 패턴.
주 사용처는 orchestra Round 2 응답 대기(idea.md Step 3.6).
1. Monitor Tool 개요
Monitor는 실행 중인 command의 stdout을 실시간 tail하여 각 라인을 notification으로
전달한다. until-loop/polling sleep 보다 저렴하다 — 대기 중 토큰 소비 없음.
기본 호출:
Monitor(
command = "...",
timeout_ms = 180000
)
반환값: notification stream (stdout의 각 라인이 1개 이벤트).
활성화 조건: AUTOPUS_PLATFORM == claude-code AND features.cc21.monitor_enabled: true.
비 OMP CLI 환경에서는 polling fallback으로 graceful degradation (R8).
2. grep --line-buffered 요구사항
Monitor가 실시간으로 라인을 수신하려면 command가 line-buffered 모드로 출력해야 한다. stdout이 pipe 앞단이면 기본값은 fully-buffered — 이벤트가 버퍼 flush 전까지 지연된다.
올바른 패턴
# GNU grep (Linux 기본값, macOS Homebrew grep)
grep --line-buffered -E 'pattern'
# BSD grep (macOS 기본값) — stdbuf로 강제 line-buffering
stdbuf -oL grep -E 'pattern'
# Universal fallback — unbuffer (expect 패키지)
unbuffer grep -E 'pattern'
잘못된 패턴 — Monitor가 실시간 이벤트를 받지 못함
grep -E 'pattern' # fully-buffered 기본값 — 이벤트 지연
cat file | grep 'pattern' # grep 자체도 line-buffered 필요
3. 플랫폼 감지 및 분기 템플릿
# Detect GNU vs BSD grep and set GREP_LB accordingly
if grep --version 2>/dev/null | grep -qi "gnu grep"; then
GREP_LB="grep --line-buffered"
else
GREP_LB="stdbuf -oL grep"
fi
Monitor 호출 시 적용:
Monitor(
command = f"cmux read-screen --surface {surface_id} --follow --scrollback 200 | {GREP_LB} -E '{idle_regex}'",
timeout_ms = 180000
)
4. Provider별 idle prompt 정규식
spec.md R7 정의. provider pane이 응답을 마치고 입력 대기 상태에 진입하면
해당 라인이 stdout으로 출력된다.
| Provider | Pattern |
|---|---|
| claude | ^❯\s*$ |
| codex | ^codex>\s*$ |
| antigravity-cli | ^>\s*(Type your|Press Ctrl) 또는 gemini>\s*$ |
| opencode | ^opencode›\s*$ 또는 ^>\s*Ready\s*$ |
오버라이드
.autopus/project/orchestra-patterns.yaml:
orchestra_patterns:
claude:
idle: '^❯\s*$'
codex:
idle: '^codex>\s*$'
antigravity-cli:
idle: '^>\s*(Type your|Press Ctrl)'
opencode:
idle: '^opencode›\s*$'
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.
- 4d ago First seen · 198 lines · 16 tokens per session scan A 540eab256ad9
monitor-patterns is a skill published in the GitHub repository Insajin/autopus-adk (110 stars, last pushed today), licensed MIT. It adds 16 tokens to every session and 1,739 once invoked, about $0.0001 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 2026-09-03.
Other skills, from other repositories
automations
Create, inspect, edit, archive, delete, run, and search scheduled or executable automation assets.
job
Execution behavior for one saved automation job running in a fresh agent conversation.
api-interface-design
Use when designing public APIs, module boundaries, provider adapters, tool schemas, or data contracts.
explore
Explore the codebase and summarize how the project is wired.
Using OmniForge Framework
Primary guide for using OmniForge CLI, MCP Server, workflows, and agent skills effectively.
github-code-review
Comprehensive GitHub code review with AI-powered swarm coordination.