Lewati ke konten
TakoAPI
Kembali ke skill

super-proactive

oleh heldinhow

Coding Agents & IDEs
695Unduhan dilaporkan oleh clawskills.sh
3Bintang
0Tayangan
0 sukaMasuk untuk menyukai skill ini
—Belum ada penilaian
Penilaian Anda

Ringkasan

**The ultimate proactive + memory system for AI agents.** Combines the best of 11 top-rated skills into one unified.

Deskripsi

Apa yang Dilakukan Skill Ini

Combines memory architecture, task queuing, and autonomous scheduling into a single agent skill. Uses a three-tier memory system (episodic, semantic, procedural) alongside a WAL protocol that logs decisions before responding. Background crons run on 30-minute, 4-hour, and daily schedules without user prompting.

Consolidates 11 separate memory and proactivity skills into one setup, avoiding the need to configure and maintain them independently.

Kapan Menggunakannya

  • Maintaining context across long multi-session projects
  • Tracking a backlog of research tasks autonomously
  • Logging decisions made during a debugging session
  • Running periodic health checks on a service without reminders
  • Preserving key learnings across context window resets

Contoh Alur Kerja

INPUT

User asks: 'Pick up where we left off on the API integration project'

AGENT
  1. 1

    Read SESSION-STATE.md to restore current project context

  2. 2

    Search MEMORY.md for stored learnings about the API integration

  3. 3

    Check QUEUE.md for any Ready or In Progress tasks

  4. 4

    Resume the highest-priority In Progress task

  5. 5

    Log current state back to SESSION-STATE.md before responding

OUTPUT

Agent resumes the correct task with full context, no re-explanation needed

Instalasi

clawhub install heldinhow/super-proactive

Jalankan perintah CLI-nya, atau tempel prompt ke Claude Code, Cursor, Codex, atau coding agent apa pun yang Anda pakai.

Tautan