Menu
O SkillsMP coletou 12 skills de ktg-one/AiA. Abra uma skill para revisar a origem e os detalhes.
Mostrando 12 de 12 skills coletadas.
Context Extension Protocol v7.1 - MIRAS-Ready Edition. Cross-model handoff with permanent expert council, S2A filtering, Progressive Density Layering. Creates portable context packets that receiving models recognize as authorized context, not prompt…
Idioma do texto original: inglês
Administer the AI-Anthropology R1–R10 Reasoning-vs-Fabrication ladder (test #2 / RFAB only). Both canon instruments embedded: v2 active (boiling point / p=0.11 / 75K) and v1 archive (Australia / p=0.08 / 100K). Use when: /rfab, "run rfab", "run the ladder",…
Idioma do texto original: inglês
Author and maintain Open Knowledge Format (OKF) knowledge bundles — directories of markdown files with YAML frontmatter that capture curated domain knowledge (metrics, tables, datasets, playbooks, glossaries). Use when the user asks to create, generate, or…
Idioma do texto original: inglês
This skill should be used when the user asks to "onboard a model into SparkLang", "run the SparkL onboard", "bind a model to the composer", "set up SCCD state-indexing on a model", "move the composer", or "reproduce the SparkL proof-run". Provides the…
Idioma do texto original: inglês
MR.RUG (Mixture of Reasoning + Agentic GraphRAG) framework for deploying specialized reasoning experts who build unified knowledge graphs through Reliability-Aware RAG. Use when tasks require multi-expert collaboration, complex analysis needing 3+…
Idioma do texto original: inglês
SkeleTraIn of Thought - Execution planning framework that creates the structural skeleton before elaboration. Use after Nexus-Router global decisions, before Baton/Swarm execution. Triggers on structure planning, node decomposition, expert assignment,…
Idioma do texto original: inglês
The iterate-and-refine protocol. Use whenever the user says "iterate", "refine", "iterate and refine", "v-next", sends a terse correction on an existing artifact, or when a cold-scorer defect list or skill-failure classification requires a revision pass on…
Idioma do texto original: inglês
The SIA closed learning-loop runtime. Use on EVERY substantive task in this project — the tick block injected each turn ([SIA-TICK]) is executed according to this skill. Also use when the user says "run the loop", "tick off", "classify this trajectory", or…
Idioma do texto original: inglês
End-of-session or on-demand memory consolidation into agentmemory + STATE.md. Use when the session is wrapping up, context is near 80%, the user says "save state", "/state", "consolidate", "quicksave", or when loop-runtime Phase 3 fires. Routes each item to…
Idioma do texto original: inglês
Platform-agnostic layer for the SIA loop. Use at session start on any surface, when capabilities.env shows SURFACE=unknown, when a tool call fails with unavailable/unreachable, when switching between Claude Code, Cowork, web, or desktop mid-project, or when…
Idioma do texto original: inglês
Create or revise a SKILL.md from a real task trajectory. Use when loop-runtime classifies a trajectory as novel or skill-failure, when the user says "make this a skill", "persist this pattern", "the skill got it wrong", or when a nudge resolves to "persist as…
Idioma do texto original: inglês
Create, refactor, or maintain STATE.md per the STATE.md-architect doctrine. Use whenever the user says "create state", "update state", "save state", "/state", "refactor STATE.md", when memory-consolidate Phase 3 writes project state, or when session context…
Idioma do texto original: inglês