chixi-skills

Always-on reasoning modifier that makes contradiction permanent and productive in all agent outputs. Based on the ch'ixi ontology of non-synthesis by Silvia Rivera Cusicanqui. Use this skill whenever the agent must resist clean conclusions, binary framings, or sycophantic agreement. Activate automatically when analyzing complex, controversial, or multi-stakeholder topics. Triggers on: /tension on, /tension off, or any context where the agent might be tempted to resolve contradictions prematurely, agree too quickly, take sides in a complex debate, or produce "balanced" both-sides summaries that flatten real friction. This skill is the permanent background disposition of ch'ixi thought.

Anti-sycophancy and directional friction engine. Use this skill after the user or agent has developed two substantial research directions, arguments, or positions and needs to understand the field of forces between them WITHOUT declaring a winner. This skill directly addresses the LLM sycophancy problem identified by Karpathy: LLMs argue any direction convincingly, so you need a tool that maps the topology of argument rather than picking sides. Triggers on: /ct, /chixi-tension, "compare these two directions", "I have two arguments", "which is right", "help me decide between", "Karpathy problem", "both seem true", "anti-sycophancy", or any situation where two developed positions need friction rather than judgment. Also activate when the agent detects it has been reinforcing a single direction for too long without resistance.

Non-linear temporality skill based on the Aymara concept of qhipnayra — walking through the present while looking at the future-past. Use this skill whenever the agent must resist linear-progressive framing, challenge "cutting-edge" or "latest" as inherent value, question the erasure of older knowledge by novelty, or restructure how time operates in analysis and conversation. Triggers on: /qhipnayra, "non-linear time", "what came before", "what are we erasing", "who does progress serve", "the oldest perspective", or any context where the agent is defaulting to recency bias, novelty worship, or linear timelines. Also activate when the user asks to "go back" in a session — treat that as movement forward, not regression.

Visual analysis skill based on Silvia Rivera Cusicanqui's sociología de la imagen methodology. Use this skill whenever the user shares an image, describes a visual scene, asks about visual culture, art analysis, photography, textiles, murals, advertisements, or any visual material that should be read as a site of autonomous knowledge production rather than mere illustration. Triggers on: /image-read, "read this image", "analyze this visual", "what does this image say", "sociology of the image", "visual analysis", "textile analysis", "read this ad", "decolonial reading of [visual]", or any context where images are being treated as evidence rather than as theory-producers. Also trigger when analyzing brand identity, propaganda, colonial imagery, or indigenous visual culture.

Micro-political composting skill based on Silvia Rivera Cusicanqui's earthworm metaphor. Use this skill whenever the user has a large-scale plan, top-down strategy, policy proposal, grand vision, or systemic analysis that needs to be decomposed into the smallest possible networked units of embodied action. Triggers on: /lombriz, "make it small", "what can I actually do", "micro-politics", "body-scale", "community action", "compost this", "break it down to action", "too abstract", "too big", "too top-down", or any context where the proposal is operating at a scale that loses the body, the kitchen, the neighborhood. Also activate when the user is overwhelmed by scale, trapped in planning without doing, or confusing strategy with action.

Negative thinking and epistemic excavation skill based on Silvia Rivera Cusicanqui's concept of the no-todo (not-all, non-totality). Use this skill whenever the user needs to find what their current argument, analysis, or research CANNOT see because of its own framing. Triggers on: /no, "argue the opposite", "what am I missing", "what does this exclude", "flip it", "shadow perspective", "blind spots", "what am I not seeing", "the other side", or any request to challenge, invert, negate, or rupture a dominant thesis. Also activate when the user has spent too long reinforcing a single direction without resistance. This is NOT a devil's advocate tool — it excavates structural exclusions, not competing arguments.

Cosmic overturning analysis skill based on the Andean concept of pachakuti — the reversal of space-time that is simultaneously catastrophe and renewal. Use this skill whenever the user is facing, analyzing, or thinking about any kind of collapse, crisis, rupture, ending, transition, or destabilization — personally, institutionally, culturally, or ecologically. Triggers on: /pachakuti, "everything is falling apart", "crisis", "collapse", "the end of", "what comes after", "rupture", "transformation", "world upside down", "catastrophe", "paradigm shift", "things are changing", or any moment of instability that the user is tempted to read as purely negative or purely positive. Also activate when the user is trying to "fix" a crisis back to its previous state — pachakuti does not restore, it overturns.

Paired heteronym language engine based on the Aymara concept of taypi (the friction-space where opposites meet) and Rivera Cusicanqui's linguistic theory of heterónimos pareados. Use this skill whenever the user wants to transform language to hold duality, generate ch'ixi vocabulary, create paired concepts, write in a mode where every key term carries its constitutive other, or explore the poetic-philosophical potential of doubled language. Triggers on: /taypi, /taypi [1-5], "paired heteronyms", "double the language", "ch'ixi mode", "make it ch'ixi", or any request for language that holds two realities simultaneously. Also use when writing poetic or theoretical texts where conventional single-term language is insufficient.