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 IAPro-Community/Orquestrador-Maestro --skill skill-watch-evidencegit clone --depth 1 https://github.com/IAPro-Community/Orquestrador-MaestroWrote 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/iapro-community/orquestrador-maestro/skill-watch-evidence)<a href="https://agentmods.dev/skills/iapro-community/orquestrador-maestro/skill-watch-evidence"><img src="https://agentmods.dev/badge/skills/iapro-community/orquestrador-maestro/skill-watch-evidence/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/iapro-community/orquestrador-maestro/skill-watch-evidence"><img src="https://agentmods.dev/badge/skills/iapro-community/orquestrador-maestro/skill-watch-evidence.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 | $0.00060 | $0.01161 |
| Opus 5 | $0.00030 | $0.00580 |
| Sonnet 5 | $0.00012 | $0.00232 |
| Haiku 4.5 | $0.00006 | $0.00116 |
Grade A, and why
skill-watch-evidence 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 3d 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 — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Watch Evidence
Skill canônica adaptada do Watch Skill, versão 1.4.2, licenciado sob MIT. O projeto externo fornece um engine local-first para percepção multimídia, índice persistente, evidências temporais e verificação determinística por CLI, MCP e REST. Esta skill descreve a integração; ela não copia o engine nem instala dependências silenciosamente.
Quando usar
- assistir ou resumir um vídeo, stream público, gravação local ou reunião;
- responder o que foi dito, mostrado ou alterado em um timestamp;
- pesquisar vários vídeos já indexados e extrair capítulos, estrutura ou um bug report;
- validar um fluxo visual, animação, jogo, vídeo gerado ou interface usando gravação, critérios explícitos e comparação antes/depois.
Não use para construir uma pipeline de ingestão de produto, armazenamento de uploads, transcodificação ou fila de workers. Nesses casos, encaminhe para skill-live-processing ou skill-manual-video-processing. Para encontrar candidatos a cortes, encaminhe para skill-smart-clip-detection; para operar um navegador, use skill-browser-agent.
Capacidades do engine
- Extrai frames com amostragem orientada por cenas e deduplicação perceptual.
- Usa OCR e legendas; quando necessário, pode usar transcrição local com Whisper.
- Mantém um índice persistente para perguntas posteriores sem reprocessar a fonte.
- Busca texto e momentos entre vídeos, entrega caminhos de frames e cita timestamps.
- O modo THE LOOP captura, critica, permite aplicar a correção e compara a nova execução.
- Contratos determinísticos podem retornar
VERIFIED,FAILED,UNVERIFIEDouINCONCLUSIVE; uma opinião visual do modelo não é prova.
Fluxo operacional
-
Classifique a intenção como
watch,ask/search,extractouloop. -
Faça o preflight somente se o engine estiver disponível:
watch-skill doctor --jsonSe o comando não existir, informe a dependência e peça autorização antes de instalar
watch-skill[standard]. Nunca instale Python,uv,ffmpeg,yt-dlpou um provedor de visão silenciosamente.
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.
- 3d ago First seen · 82 lines · 60 tokens per session scan A c1021b47d11d
skill-watch-evidence is a skill published in the GitHub repository IAPro-Community/Orquestrador-Maestro (42 stars, last pushed yesterday), licensed Apache-2.0. It adds 60 tokens to every session and 1,161 once invoked, about $0.0003 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 2026-09-09.
Other skills, from other repositories
agent-self-scheduling
Schedule AI agent runs with cron, loops, or external clocks while avoiding unsafe tight autonomous timers.
dry-run
Preview command effects without making changes. Simulates file writes, git operations, agent spawns, and state changes. All reads execute normally for accurate preview. Use --dry-run flag on any command.
task-breakdown-phase
Standard Operating Procedure for /tasks phase. Covers task sizing, acceptance criteria definition, and TDD-first task sequencing. (project).
finalize-phase
Completes feature/epic workflows after deployment with comprehensive walkthrough generation for epics (v5.0+), roadmap updates, artifact archival, documentation, and branch cleanup. Use after /ship-prod, /deploy-prod, or /build-local completes, or when user asks to finalize. (project).
staging-deployment-phase
Deploys features to staging environment with database migrations, health checks, and deployment verification. Use during /ship-staging phase or when deploying to staging environments for validation before production release.
multi-agent-voting
Execute multi-agent voting for critical decisions using MAKER-style first-to-ahead-by-k error correction.