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 BoxLogoDev/sapstack --skill sap-sessiongit clone --depth 1 https://github.com/BoxLogoDev/sapstackWrote 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/boxlogodev/sapstack/sap-session)<a href="https://agentmods.dev/skills/boxlogodev/sapstack/sap-session"><img src="https://agentmods.dev/badge/skills/boxlogodev/sapstack/sap-session/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/boxlogodev/sapstack/sap-session"><img src="https://agentmods.dev/badge/skills/boxlogodev/sapstack/sap-session.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00121 | $0.02663 |
| Opus 5.5 | $0.00048 | $0.01065 |
| Sonnet 5.5 | $0.00024 | $0.00533 |
| Haiku 4.5 | $0.00012 | $0.00266 |
Grade A, and why
sap-session 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 — 201 lines — stays where its author put it; the contents beside it link to each section on GitHub.
sap-session — Evidence Loop Orchestrator
🎯 이 스킬의 존재 이유
대기업 SAP 환경에서는 라이브 시스템 접근이 금지됩니다. 그래서 AI는 실행기가 될 수 없고, 조언자에 머물 수밖에 없습니다. 이 제약을 제거하는 대신 수용하면서도 "조언봇"을 넘어서는 방법이 Evidence Loop입니다:
- 운영자가 실행기 역할을 맡아 SAP에서 증거를 가져오고,
- AI는 각 턴마다 가설·요청·검증·판정을 담당하며,
- 세션 상태는 파일에 저장되어 중단·재개·인수인계가 가능합니다.
즉, Human-in-the-loop 비동기 루프입니다. 각 턴 사이에 몇 분·몇 시간·며칠이 걸려도 좋고, 다른 운영자가 이어받아도 맥락이 유지됩니다.
🔁 네 개의 턴
┌── Turn 1 · INTAKE ───────────────────────────────────┐
│ 운영자/엔드유저가 초기 증상 + 최소 증거를 업로드 │
│ 산출: Evidence Bundle 1개 │
│ 세션 상태: intake → hypothesizing │
└──────────────────────────────────────────────────────┘
↓
┌── Turn 2 · HYPOTHESIS ───────────────────────────────┐
│ AI가 2~4개 가설 + 각 가설의 반증 조건 + Follow-up │
│ Request(운영자가 다음 수집할 체크리스트) 생성 │
│ 산출: hypotheses[] + followup_request 1개 │
│ 세션 상태: hypothesizing → awaiting_evidence │
└──────────────────────────────────────────────────────┘
↓
┌── Turn 3 · COLLECT ──────────────────────────────────┐
│ 운영자가 Follow-up Request의 체크리스트를 SAP에서 │
│ 실행하여 새 Evidence Bundle 업로드 │
│ 산출: Evidence Bundle 1개 추가 │
│ 세션 상태: awaiting_evidence → verifying │
└──────────────────────────────────────────────────────┘
↓
┌── Turn 4 · VERIFY ───────────────────────────────────┐
│ AI가 새 증거로 가설을 확정/기각. 확정 시 Fix + 필수 │
│ Rollback 플랜. 기각 시 새 가설 → Turn 2 재진입. │
│ 산출: verdict 1개 │
│ 세션 상태: verifying → (resolved | hypothesizing) │
└──────────────────────────────────────────────────────┘
세부 포맷과 예시는 references/turn-formats.md에 있습니다.
📦 데이터 모델
모든 턴의 입출력은 schemas/ 아래 5개 YAML JSON Schema로 정의됩니다:
| 스키마 | 언제 생성되는가 | 어디에 저장되는가 |
|---|---|---|
evidence-bundle.schema.yaml |
Turn 1, Turn 3 | .sapstack/sessions/{id}/bundles/evb-*.yaml |
hypothesis.schema.yaml |
Turn 2 | .sapstack/sessions/{id}/hypotheses/h-*.yaml |
followup-request.schema.yaml |
Turn 2 | .sapstack/sessions/{id}/requests/flr-*.yaml |
verdict.schema.yaml |
Turn 4 | .sapstack/sessions/{id}/verdicts/vdc-*.yaml |
session-state.schema.yaml |
매 턴 업데이트 | .sapstack/sessions/{id}/state.yaml |
What ships with it
11 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/de/quick-guide-de.md 4.7 KB
- references/en/quick-guide-en.md 4.6 KB
- references/evidence-bundle-guide.md 7.2 KB
- references/followup-authoring.md 9.0 KB
- references/ja/quick-guide-ja.md 4.9 KB
- references/ko/quick-guide.md 7.0 KB
- references/korean-field-language.md 9.2 KB
- references/session-state-lifecycle.md 9.6 KB
- references/turn-formats.md 9.8 KB
- references/vi/quick-guide-vi.md 5.1 KB
- references/zh/quick-guide-zh.md 4.3 KB
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 · 201 lines · 121 tokens per session scan A 43d3ae7065b0
sap-session is a skill published in the GitHub repository BoxLogoDev/sapstack (20 stars, last pushed 26d ago), licensed MIT. It adds 121 tokens to every session and 2,663 once invoked, about $0.0005 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 2026-09-04.
Other skills, from other repositories
systemic-issue-triage
Trigger: new issue, bug report, triage, backlog, issue flood, community report, root cause, dead-end, blocked user. Attack issues by root class, never one-by-one; fixes must shrink the system, not grow it.
triage-issue-reports
Triage Slack issue reports with one thread-only verdict, evidence review, cause-aware routing, tracker dedupe, and fail-closed ticket creation. Use only from the configured Benny triage automation.
sap-background-jobs
Investigate SAP background jobs, including failed or aborted jobs, job status, job steps, job logs, runtime dumps, TBTCO, TBTCP, and TBTCJOBLOG0-9.
abap-performance-ecc
ABAP performance best practices for ECC / traditional database systems (Oracle, DB2,MSSQL, MaxDB). Use when writing or reviewing ABAP code on NON-HANA systems.IMPORTANT: First use the SAP system info tool to check the system type — if the system is S/4HANA or runs on HANA DB, load the abap-performance-hana skill…
om-auto-fix-issue
Fix or implement a tracker issue end to end from a single command — takes an issue id or a plain problem description (filed first via om-prepare-issue), classifies, then drives the bug autofix chain (om-verify-in-repo, om-root-cause, om-fix, om-open-pr, om-auto-review-pr, om-auto-qa-pr for UI fixes) or the feature…
factory-triage
Triage a Factory work item's issue — trace history, understand architecture, diagnose root cause, then advance the stage.