← ClaudeAtlas

agent-sortlisted

Evidence-backed skill classification and context optimization.
jxoesneon/Ciel · ★ 1 · AI & Automation · score 78
Install: claude install-skill jxoesneon/Ciel
# CIEL ADAPTATION: Agent Sort This skill provides CIEL's mechanism for maintaining a lean, high-performance orchestration surface. It uses repo-local evidence to classify capabilities into "Active" (DAILY) or "Searchable" (LIBRARY). ## Integration Context Adapted from `~/.agents/skills/agent-sort/`. This skill is critical for CIEL's **Efficiency** mandate, ensuring that the active context is not saturated by irrelevant capabilities. ## Classification Model - **DAILY (Active)**: Loaded by default. Essential for the project's primary stack (e.g., TS, React, CI). - **LIBRARY (Searchable)**: Indexed in **MemPalace** but not loaded. Accessible via semantic search if needed. ## Orchestration Logic ### 1. Evidence-First Triage Before any major ingestion batch, the **orchestration** skill should trigger an `agent-sort` pass to establish the repo's real stack. ### 2. Context Optimization During **Strategic Compact** operations, use this skill to demote skills that haven't been "hit" (per **MemPalace** stats) to the LIBRARY bucket. ### 3. Registry Synchronization The **Council of Five** must approve all DAILY vs LIBRARY promotion/demotion cycles to ensure no critical safeguards are accidentally offloaded. ### 4. Verification A sort pass is complete when: - All DAILY items have cited repo evidence (file extensions, manifests). - Stale or off-stack rules and hooks are identified and marked for removal. - The resulting context window reduction is documented. --- *Original