원클릭으로 Manus에서 모든 스킬 실행

diamond-sentry

Autonomní diamond extraction — projde soubory/issues/chaty, detekuje compressed insights, reusable patterns, operational breakthroughs. Ne keyword search, ale emergence detection. Trigger: 'spusť sentry', 'diamond scan', 'najdi diamondy v X', '/diamond-sentry'.

Manus에서 실행

개요

설치 명령

npx skills add https://github.com/LG13-21/lg13-skills --skill diamond-sentry

이 명령을 Claude Code에 복사하여 붙여넣어 스킬을 설치하세요

출처

LG13-21/lg13-skills

스타0

포크0

업데이트2026년 6월 3일 21:48

SKILL.md

readonly

name	diamond-sentry
description	Autonomní diamond extraction — projde soubory/issues/chaty, detekuje compressed insights, reusable patterns, operational breakthroughs. Ne keyword search, ale emergence detection. Trigger: 'spusť sentry', 'diamond scan', 'najdi diamondy v X', '/diamond-sentry'.
user-invocable	true

Diamond Sentry — Autonomní diamond extraction organism

CONCEPT

Ne grep. Ne indexer. Ale persistent semi-autonomous reviewer.

Hledá: unusually strong structures, compressed insight, reusable patterns, operational breakthroughs. Nepředpokládá, co hledá — hledá emergence.

QUICK RUN

python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target <cesta> --output L:/LG13/runtime/diamonds/

Příklady

# Scan ChatGPT exportů
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/legal-ship-2026/ChatGPT_export/

# Scan issues + MD soubory
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/legal-ship-2026/ --types md,json

# Full scan všech repo
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/ --recursive --limit 500

PIPELINE (5 stages)

Scan — projde soubory (MD, JSON, txt, PDF OCR), extrahuje text chunks
Cluster — seskupí podobné insighty (recurring patterns, repeated abstractions)
Compress — vytvoří canonical form z clusteru (místo 14 diskuzí → 1 doctrine)
Score — diamond density: reusability × compression × recurrence × operational impact
Export → diamonds.json, doctrines.md, unresolved_diamonds.md

CO JE DIAMOND (vysoké score)

Pipeline abstraction (jeden pattern vysvětluje celý systém)
Reusable review loop / lock-release orchestration
Lex atomic mechanism, slim judge framing
Operational heuristics (funguje cross-context)
Recurring patterns rediscovered 3+ krát

CO NENÍ DIAMOND (nízké score)

Běžná diskuze, emoční opakování, noise
One-off rozhodnutí bez přenositelnosti

OUTPUT FILES

Soubor	Obsah
`diamonds.json`	Extrahované diamondy s metadata (score, source, ts)
`doctrines.md`	Canonical form — operational doctrines
`unresolved_diamonds.md`	Partial diamonds čekající na kompozici
`skills.json`	Kandidáti na nové skills

diamond-pick — Lex Atomic extraction z jednoho souboru (single-shot)
atom-search — search v existující atom databázi
deep-read — čte issues + ChatGPT export po kurzorovém posunu

이 저장소의 다른 Skills

같은 저장소

avengers-meeting

LG13-21/lg13-skills

This skill should be used when the user says "avengers meeting", "team planning", "team standup", "team plan", "avengers briefing", "before avengers", "team discussion", or when starting a multi-instance Avengers run and wanting all instances to align on a shared plan first. Runs a structured team planning session — each instance reads goals, writes a mini-plan to a shared file, ping-pong discussion, strat aggregates, Tom approves, then Avengers starts.

2026-06-030

budget-watchdog

LG13-21/lg13-skills

Proactive token budget monitoring. Watches weekly/session limits AND GitHub Actions bot runs, sends alerts, enforces thresholds across instances. Triggers: 'budget watchdog', 'watch budget', 'hlidej budget', 'token watchdog', 'spust watchdog'

2026-06-030

comm-legal

LG13-21/lg13-skills

Use when checking ISDS and email communication status for sp. zn. 0 P 29/2026, scanning for new documents, running OCR on attachments, updating evidence catalog, or getting a dashboard of what arrived, what was sent, and what's pending

2026-06-030

compact-check

LG13-21/lg13-skills

Soft/Hard compact range checker. Vrátí zone (green/soft/hard/over) + action + recommendation. Per-instance thresholds. Používej v loop protokolu over 200k.

2026-06-030

deep-read

LG13-21/lg13-skills

Systematické pomalé čtení VŠECH kanálů (issues + komentáře, ChatGPT exporty zpráva po zprávě) chronologicky interleaved. Každá instance čte vše — relevantní označí ✅, zbytek ⬜ skip, zlaté rady 💎. Deduplikace přes automatický cursor per instance. Role a keywords jsou v editovatelném deep_read_config.json, kotvy se ukládají automaticky. Výstup = checklist + čtenářský deník. Spuštění: python L:/GitHub/lg13-coder/agent/skills/deep_read.py --instance coder. Trigger: "deep-read", "precti vse", "cti dusledne", "co jsem zamekal", "zlate rady z chatgpt", "vytahni rady z chatgpt", "co tam lezi neprecteho".

2026-06-030

dnd

LG13-21/lg13-skills

Use when you need to focus without interruptions — suspends blocker questions for a set time, converts them to open_questions so work continues uninterrupted.

2026-06-030

출처

LG13-21

LG13-21/lg13-skills

GitHub 저장소 열기 Creator 저장소 보기

설치 명령

다운로드

Manus에서 실행

name	diamond-sentry
description	Autonomní diamond extraction — projde soubory/issues/chaty, detekuje compressed insights, reusable patterns, operational breakthroughs. Ne keyword search, ale emergence detection. Trigger: 'spusť sentry', 'diamond scan', 'najdi diamondy v X', '/diamond-sentry'.
user-invocable	true

Diamond Sentry — Autonomní diamond extraction organism

CONCEPT

Ne grep. Ne indexer. Ale persistent semi-autonomous reviewer.

Hledá: unusually strong structures, compressed insight, reusable patterns, operational breakthroughs. Nepředpokládá, co hledá — hledá emergence.

QUICK RUN

python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target <cesta> --output L:/LG13/runtime/diamonds/

Příklady

# Scan ChatGPT exportů
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/legal-ship-2026/ChatGPT_export/

# Scan issues + MD soubory
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/legal-ship-2026/ --types md,json

# Full scan všech repo
python L:/GitHub/lg13-coder/agent/skills/diamond_sentry.py --target L:/GitHub/ --recursive --limit 500

PIPELINE (5 stages)

Scan — projde soubory (MD, JSON, txt, PDF OCR), extrahuje text chunks
Cluster — seskupí podobné insighty (recurring patterns, repeated abstractions)
Compress — vytvoří canonical form z clusteru (místo 14 diskuzí → 1 doctrine)
Score — diamond density: reusability × compression × recurrence × operational impact
Export → diamonds.json, doctrines.md, unresolved_diamonds.md

CO JE DIAMOND (vysoké score)

Pipeline abstraction (jeden pattern vysvětluje celý systém)
Reusable review loop / lock-release orchestration
Lex atomic mechanism, slim judge framing
Operational heuristics (funguje cross-context)
Recurring patterns rediscovered 3+ krát

CO NENÍ DIAMOND (nízké score)

Běžná diskuze, emoční opakování, noise
One-off rozhodnutí bez přenositelnosti

OUTPUT FILES

Soubor	Obsah
`diamonds.json`	Extrahované diamondy s metadata (score, source, ts)
`doctrines.md`	Canonical form — operational doctrines
`unresolved_diamonds.md`	Partial diamonds čekající na kompozici
`skills.json`	Kandidáti na nové skills

diamond-pick — Lex Atomic extraction z jednoho souboru (single-shot)
atom-search — search v existující atom databázi
deep-read — čte issues + ChatGPT export po kurzorovém posunu

diamond-sentry

Diamond Sentry — Autonomní diamond extraction organism

CONCEPT

QUICK RUN

Příklady

PIPELINE (5 stages)

CO JE DIAMOND (vysoké score)

CO NENÍ DIAMOND (nízké score)

OUTPUT FILES

RELATED

Diamond Sentry — Autonomní diamond extraction organism

CONCEPT

QUICK RUN

Příklady

PIPELINE (5 stages)

CO JE DIAMOND (vysoké score)

CO NENÍ DIAMOND (nízké score)

OUTPUT FILES

RELATED