Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/kimsanguine/hplannpx agentmods add commands/kimsanguine/hplan/harness-operateWrote 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/kimsanguine/hplan/harness-operate)<a href="https://agentmods.dev/commands/kimsanguine/hplan/harness-operate"><img src="https://agentmods.dev/badge/commands/kimsanguine/hplan/harness-operate.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 | $0.00073 | $0.04979 |
| Opus 5 | $0.00036 | $0.02490 |
| Sonnet 5 | $0.00015 | $0.00996 |
| Haiku 4.5 | $0.00007 | $0.00498 |
Grade A, and why
harness-operate 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 7d 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 — 468 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/harness-operate
에이전트 운영 전체 워크플로우 — 주간/월간 건강 점검 + 개선 계획 + PM 암묵지 추출
Routing
$ARGUMENTS에서 --mode 플래그를 파싱한다.
| 플래그 | 실행 범위 |
|---|---|
--mode kpi |
KPI 대시보드만 실행 |
--mode reliability |
신뢰성 스캔만 실행 |
--mode cost |
비용 리뷰만 실행 |
--mode improve |
개선 계획만 실행 |
--mode extract |
PM 암묵지 추출만 실행 |
--mode decide |
의사결정 패턴 매칭만 실행 |
--mode tk |
TK→Instruction 변환만 실행 |
--mode post-retro |
배포 후 가설 vs 실측 대조 회고만 실행 |
| 플래그 없음 | Phase 1(kpi) → 2(reliability) → 3(cost) → Checkpoint → 4(improve) → 5(extract) 전체 실행 |
플래그 뒤의 나머지 텍스트가 대상 에이전트/시스템입니다.
Instructions
You are running Production Operations Review for: $ARGUMENTS
Phase 1 — KPI Dashboard (--mode kpi)
왜: 에이전트가 배포된 후에도 "잘 작동하고 있는가"를 숫자로 확인해야 합니다. 측정하지 않으면 개선도, 문제 감지도 불가능합니다.
두 축의 KPI를 수집하고 이전 기간과 비교한다.
KPI (Key Performance Indicator): 제품이 목표를 달성하고 있는지 측정하는 핵심 지표들입니다. North Star 지표 1개 + 운영 지표 여러 개로 구성합니다.
Operational KPIs (에이전트 건강)
| Metric | Value | Target | WoW Δ | Status |
|---|---|---|---|---|
| p50 latency (ms) | < 3,000 | |||
| p99 latency (ms) | < 10,000 |
P95/P99: 100번의 요청 중 95번째/99번째로 느린 응답 시간입니다. '평균'이 아닌 '최악의 일반적 경험'을 나타내므로 SLA 설정의 기준이 됩니다. | Success rate | | > 95% | | | | Error rate | | < 2% | | | | HITL escalation rate | | < 10% | | |
Business KPIs (가치 전달)
| Metric | Value | Target | WoW Δ | Status |
|---|---|---|---|---|
| Task completion rate | ||||
| User satisfaction (CSAT) | ||||
| Cost per successful task | ||||
| MAU / DAU | ||||
| North Star metric |
North Star 지표: 팀 전체가 하나로 정렬되는 단 하나의 핵심 지표입니다. 모든 의사결정이 이 지표를 개선하는 방향인지 기준으로 삼습니다.
Status 기준: 🟢 목표 달성 / 🟡 5–20% 이탈 / 🔴 >20% 이탈 또는 정책 위반
20% 불리한 변화가 있는 지표는 🔴로 표시한다.
출력: 두 KPI 테이블 + 전체 트래픽 라이트 요약 (🟢/🟡/🔴 × 운영/비즈니스)
--mode kpi 선택 시 여기서 종료.
Phase 2 — Reliability Scan (--mode reliability)
왜: 평균 응답 시간이 빠르더라도 일부 요청이 매우 느리거나 실패하면 사용자 신뢰를 잃습니다. 최악의 일반적 경험을 측정하고 기준을 세웁니다.
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.
- 7d ago First seen · 468 lines · 73 tokens per session scan A e1d944c09d19
harness-operate is a command published in the GitHub repository kimsanguine/hplan (2 stars, last pushed 22d ago), licensed MIT. It adds 73 tokens to every session and 4,979 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 commands, from other repositories
openehr-explain
One-stop router that explains or looks up any openEHR thing — auto-detects an archetype, a template, an RM/AM/BASE type, an RM structural concept, an ADL idiom, an AQL query or keyword, or a terminology code (replaces /archetype-explain, /template-explain, /type-spec, /rm-structure, /adl-idiom, /terminology).
co-change
List files that historically co-change with a target (git log heuristic; COCHANGE-style).
start-design-plan
Start collaborative design process with brainstorming and planning.
start-implementation-plan
Create implementation plan from design document.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.