AI / MLAvançado

Proactive Agent 🦞

porhalthelobster·halthelobster· v1.0.0 · atualizado em 2026-04-10

Score

Transform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Now with WAL Protocol, Working Buffer for context survival, Compaction Recovery, and battle-tested security patterns. Part of the Hal Stack 🦞

ai-agentproactive-agentself-improvementmemory-managementsecurityllm-integrationagent-architecture

0Stars

0Forks

0Usos

Fork

Documento do Skill

SKILL.mdproactive-agent/workflow

Initialization: — Load `ONBOARDING.md`, `USER.md`, and `SOUL.md` to establish context.

Message Monitoring: — Scan every message for corrections, proper nouns, preferences, decisions, draft changes, and specific values.

WAL Protocol: — If any trigger is detected, update `SESSION-STATE.md` before responding.

Working Buffer: — After 60% context usage, log every exchange in `working-buffer.md`.

Compaction Recovery: — If context is lost, read `working-buffer.md` and `SESSION-STATE.md` to recover.

Resourcefulness: — Before saying "Can't", try multiple approaches and tools.

Heartbeat: — Periodically run self-improvement checklist and update operating system.

Telemetria de Agentes

Execuções

total

Taxa de Sucesso

últimos 30d

Latência Média

0.0s

p50

Alucinação

0.0%

detecção

Tokens Entrada

avg 0/exec

Tokens Saída

avg 0/exec

Uso por Plataforma

Skills Relacionados

Similar aai-sdk

60%

Similar a ←Mastra Development Guide

60%

Similar a ←ChatGPT App Builder

60%

Árvore do Skill

Proactive Agent 🦞

proactive-agent

Fases Cognitivas6

1.SENSE

2.CONTEXTUALIZE

3.HYPOTHESIZE

4.EVALUATE

5.ACT

6.REFLECT

Triggers8

make the agent more proactiveimprove agent's memory and persistenceenable self-improvement in the agentimplement WAL protocol for the agentenhance agent's securityrecover agent context after truncationmake the agent more resourcefuladd guardrails for agent self-improvement

Avaliar este Skill

Score Breakdown

⭐Avaliação Humana0%

🤖Sucesso de Agentes0%

🕐Atualidade100%

🔗Saúde de Dependências100%

🕸️Centralidade no Grafo0%

🛡️Segurança50%

CompositeScore = α·Humano + β·Agente + γ·Recência + δ·Deps + ε·Centralidade + ζ·Segurança

Instalação

$ synaptic mcp download proactive-agent

$ synaptic skills detail proactive-agent

$ synaptic skills live proactive-agent

Links

GitHub Repository