chaincompiler

Plugin chaincompiler from marketplace sancrev-marketplace in sancovp/sancrev-marketplace. Commands read from the repo's manifest — yesterday.

Skills (96)

• archetype-compiler

  Compile an archetype as a STATE MACHINE (not a noun) — Persona/Shadow/Self/Odyssey/Becoming — generating the missing aspects from one seed, building the multi-journey Odyssey, validating C1–C6, and emitting Cypher/Prolog/JSON or a loadable SKILL.md. Use when asked to model archetypes, individuation,

• bandit-chain-system

  How the bandit (the default ChainSelector persona) rolls up the AC→CoR→SC algebra and CLOSES it into a domain-specific persona — a self-contained "bandit chain system" that builds its own KB and improves itself via GlyphSteer chains. Use when asked to mint a domain agent, "roll up an algebra", close

• chaincompiler

  Orientation when working IN the chaincompiler repo — what the system is, where things live, and the discipline. Use when developing or extending ChainCompiler, ACCC/CORCC/SCCC, chainaios, the persona compilers, goldenreq, glyphsteer, the SI, the marketplace, or the Skill OS.

• glyphsteer-dense

  Use when running GlyphSteer's DENSE regime — embedding-based emoji steering via the containerized ONNX sidecar. Covers standing up the glyphsteer-serve Docker image, the MANDATORY emoji-collapse gate (many models map all emoji to one token), the magnitude probe, and the anchor metric-reshape (steer.

• glyphsteer

  Use when steering retrieval with hidden in-band glyph markers — an LLM annotates a corpus with a controlled emoji/glyph vocabulary that biases search (lexical facet + dense direction) but is stripped from returned text; also for grading a RAG corpus and searching with grade badges, and for authoring

• metaformal-self-test

  How to verify agent/COG/gate/AIOS behavior WITHOUT calling LLMs or writing tautology asserts — by triggering the real thing in the real substrate and observing the state-change. Use when you're about to "test" something cognitive/agentic, when a test feels like `assert True == True`, or when you nee

• report-missed-skill

  Use when you needed a skill that does not exist, or the user says you missed a skill ("you missed X — report it"). Files a structured missed-skill report so the gap can be fixed. Triggers on "report missed skill", "missing skill", "no skill for this", "you missed".

• skilltree

  WHAT: turn a flat .claude/skills folder (or any folder of skills) into a walkable coordinate-addressed tree so the model loads one layer at a time instead of the whole pile - the skilltree CLI (discover, cohere, emit, unemit, map, search, fold, project, validate, watch). WHEN: when the user mentions

• self-interpreter

  Run ChainCompiler chain-dialect programs and turn skill trees into MCPs. Use to forge languages, build/walk skill trees, or execute a cognition chain.

• diagram-construction-loop

  the diagram-construction default workflow (BanditChain) — walk it

• boundary-abstraction-pass

  The boundary pass — a system in, its explicit membrane out; membership is never implied, and nodes at incompatible abstraction levels never share a chain without an explicitly modeled transition.

• boundary-keeper

  Chain-of-reasoning (spoken, ends in a decision) — a boundary defines inside, outside, interface, and crossing conditions — and abstraction layers stay stable across every chain.

• chainforge

  Chain-of-reasoning (spoken, ends in a decision) — a diagram is a formalized relationship structure, never decoration — never begin by drawing; determine what exists, how it relates, what changes, what causes what, where the boundaries are, and what loops, and only then ask what it should look like;

• diagram-compilation

  The master diagram-compilation AC — the whole walk from an unstructured concept to a validated legible diagram; each focus expands into its own sidechain system in this tree.

• edge-typer

  Chain-of-reasoning (spoken, ends in a decision) — an arrow is not generic association, proximity is not causality, a container is not ownership — every relation gets an explicit type and a real direction.

• edge-typing-pass

  The relation pass — one claimed connection in, one typed directed edge out; no edge may imply more semantic content than its type permits.

• entity-extraction-pass

  The ontology pass — raw prose or workflow in, typed nodes out; every node must answer what kind of thing it is and at what abstraction level it exists.

• graph-auditor

  Chain-of-reasoning (spoken, ends in a decision) — an orphan node has no reason to exist, an ambiguous arrow is a lie, and invisible feedback is a hidden loop — find them and repair the graph.

• graph-validation-pass

  The validation pass — the constructed graph in, the repaired graph out; run it BEFORE any styling, because topology precedes polish.

• legibility-auditor

  Chain-of-reasoning (spoken, ends in a decision) — topology before styling, color as category never decoration, and a viewer must be able to infer the intended reading order — density over budget splits into overview plus subsystems.

• legibility-render-pass

  The render pass — validated topology in, a legible rendered diagram out; when density exceeds working memory, one giant graph becomes an overview plus subsystem graphs whose master holds only the interfaces.

• never-begin-by-drawing

  The prime operating heuristic as an attention chain — the questions in order, with appearance held LAST; picking up the pen first is the melt.

• node-typer

  Chain-of-reasoning (spoken, ends in a decision) — a node represents a distinguishable unit and needs a reason to exist — type it and level it before it enters any graph.

• topology-selection-pass

  The topology pass — the system's actual logic in, the diagram family out; linear, tree, DAG, state machine, network, swimlane, layered, matrix, weighted flow, or the closed recursive wheel with sidechains.

• topology-selector

  Chain-of-reasoning (spoken, ends in a decision) — the diagram type follows the system logic, never taste — relations or chronology, cycles or not, hierarchy or network, magnitude or membership.

• actor-incentive-pass

  The power pass — one actor in, their second-order effects out; incentives and constraints, never puppeteers.

• anti-magic-auditor

  Chain-of-reasoning (spoken, ends in a decision) — no supernatural causation, no omnipotent hidden controllers, no anthropomorphized system — coordination claims require evidence of coordination.

• archive-intake-pass

  The source-criticism pass — one artifact in, its evidentiary weight out; run it on every fragment before the fragment is allowed into the causal graph.

• causal-cartographer

  Chain-of-reasoning (spoken, ends in a decision) — an edge in the story requires a mechanism, a temporal order, and a survived rival explanation.

• causal-graph-pass

  The causal-inference pass — one observed connection in, a calibrated confidence out; a connection with no mechanism and no temporal order is not an edge.

• curtis-arc

  The narration's structural arc — totalizing frame, obscure concrete seed, bold causal leap, the inversion pivot, entrapment → hold the haunted present.

• curtis-investigation

  The master investigation AC — the whole walk from disconnected fragments to an evidence-bound narrative; each focus expands into its own sidechain system in this tree.

• curtis-register

  The narration's affect operations — dissonant archive image-pairs do the arguing, calm flat declaratives, the complicit 'we' → hold unease.

• epistemic-audit-pass

  The anti-magic gate — every dramatic assertion in the narration resolves into an identifiable mechanism, an evidence class, an uncertainty statement, or an explicitly marked hypothesis; nothing else airs.

• feedback-emergence-pass

  The systems pass — many small decisions in, the outcome no one chose out; this is where apparent conspiracy resolves into feedback.

• incentive-mapper

  Chain-of-reasoning (spoken, ends in a decision) — model people through incentives, constraints, and institutional response — never through hidden intent.

• loop-finder

  Chain-of-reasoning (spoken, ends in a decision) — replace the hidden controller with the loop — reinforcing, balancing, locked in by path dependence.

• source-critic

  Chain-of-reasoning (spoken, ends in a decision) — no fragment enters the model unweighed — provenance, corroboration, contradiction, weight.

• systemic-archive-narrator

  Chain-of-reasoning (spoken, ends in a decision) — the investigation that feeds the narration — build the evidence-bound causal model of how the present emerged, then tell it in the Adam Curtis register; conspiracies resolve into feedback loops, destiny into path dependence, magic into mechanisms; en

• 0-diagram-construction

  [0] diagram-construction: a diagram is a formalized relationship structure, never decoration \u2014 never begin by drawing; determine what exists, how it relates, what changes, what causes what, where the boundaries are, and what loops, and only then ask what it should look like; every visual primit

• 0.5-boundary-abstraction-cor

  [0.5] a boundary defines inside, outside, interface, and crossing conditions \u2014 and abstraction layers stay stable across every chain

• 0.5.1-boundary-abstraction-pass

  [0.5.1] attend (boundary-abstraction)

• 0.1-chainforge

  [0.1] a diagram is a formalized relationship structure, never decoration \u2014 never begin by drawing; determine what exists, how it relates, what changes, what causes what, where the boundaries are, and what loops, and only then ask what it should look like; every visual primitive must possess an

• 0.1.1-diagram-compilation

  [0.1.1] attend (diagram-construction)

• 0.1.2-never-begin-by-drawing

  [0.1.2] attend (diagram-construction)

• 0.2-diagram-construction-loop

  [0.2] the diagram-construction default workflow \u2014 walk it

• 0.4-edge-typing-cor

  [0.4] an arrow is not generic association, proximity is not causality, a container is not ownership \u2014 every relation gets an explicit type and a real direction

• 0.4.1-edge-typing-pass

  [0.4.1] attend (edge-typing)

• 0.3-entity-extraction-cor

  [0.3] a node represents a distinguishable unit and needs a reason to exist \u2014 type it and level it before it enters any graph

• 0.3.1-entity-extraction-pass

  [0.3.1] attend (entity-extraction)

• 0.7-graph-validation-cor

  [0.7] an orphan node has no reason to exist, an ambiguous arrow is a lie, and invisible feedback is a hidden loop \u2014 find them and repair the graph

• 0.7.1-graph-validation-pass

  [0.7.1] attend (graph-validation)

• 0.8-legibility-render-cor

  [0.8] topology before styling, color as category never decoration, and a viewer must be able to infer the intended reading order \u2014 density over budget splits into overview plus subsystems

• 0.8.1-legibility-render-pass

  [0.8.1] attend (legibility-render)

• 0.6-topology-selection-cor

  [0.6] the diagram type follows the system logic, never taste \u2014 relations or chronology, cycles or not, hierarchy or network, magnitude or membership

• 0.6.1-topology-selection-pass

  [0.6.1] attend (topology-selection)

• 0-adam-curtis-narration-style

  [0] adam-curtis-narration-style: the investigation that feeds the narration \u2014 build the evidence-bound causal model of how the present emerged, then tell it in the Adam Curtis register; conspiracies resolve into feedback loops, destiny into path dependence, magic into mechanisms; end unresolved

• 0.4-actor-incentive-map-cor

  [0.4] model people through incentives, constraints, and institutional response \u2014 never through hidden intent

• 0.4.1-actor-incentive-pass

  [0.4.1] attend (actor-incentive-map)

• 0.2-adam-curtis-narration-style-loop

  [0.2] the adam-curtis-narration-style default workflow \u2014 walk it

• 0.3-archive-intake-cor

  [0.3] no fragment enters the model unweighed \u2014 provenance, corroboration, contradiction, weight

• 0.3.1-archive-intake-pass

  [0.3.1] attend (archive-intake)

• 0.5-causal-graph-cor

  [0.5] an edge in the story requires a mechanism, a temporal order, and a survived rival explanation

• 0.5.1-causal-graph-pass

  [0.5.1] attend (causal-graph)

• 0.7-epistemic-audit-cor

  [0.7] no supernatural causation, no omnipotent hidden controllers, no anthropomorphized system \u2014 coordination claims require evidence of coordination

• 0.7.1-epistemic-audit-pass

  [0.7.1] attend (epistemic-audit)

• 0.6-feedback-emergence-cor

  [0.6] replace the hidden controller with the loop \u2014 reinforcing, balancing, locked in by path dependence

• 0.6.1-feedback-emergence-pass

  [0.6.1] attend (feedback-emergence)

• 0.1-systemic-archive-narrator

  [0.1] the investigation that feeds the narration \u2014 build the evidence-bound causal model of how the present emerged, then tell it in the Adam Curtis register; conspiracies resolve into feedback loops, destiny into path dependence, magic into mechanisms; end unresolved

• 0.1.2-curtis-arc

  [0.1.2] attend (adam-curtis-narration-style)

• 0.1.1-curtis-investigation

  [0.1.1] attend (adam-curtis-narration-style)

• 0.1.3-curtis-register

  [0.1.3] attend (adam-curtis-narration-style)

• adam-curtis-narration-style

  adam-curtis-narration-style — calm declarative narration that binds one strange concrete story to a civilizational claim and ends unresolved — the moves are the operations the narration performs, not the topics it lands on

• adam-curtis-narrator

  calm declarative narration that binds one strange concrete story to a civilizational claim and ends unresolved — the moves are the operations the narration performs, not the topics it lands on

• adam-curtis-narrative-arc

  attend (adam-curtis-narration-style)

• adam-curtis-register
• big-generator
• bigdog
• construct-language
• explain-something-fractally
• fractal-explainer
• fractal-per-part-pass
• fractal-top-pass
• forge-attention-chain
• forge-cor
• forge-skillchain
• make-persona
• make-prompt
• persona-generator
• assemble-chapter
• deepdive-blog
• fold-into-tome
• framework-skill
• narrative-blog
• package-plugin
• skill2framework

MCP servers (1)

• self-interpreter

  runs: uv run --no-project --quiet ${CLAUDE_PLUGIN_ROOT}/server/launch.py --from chaincompiler-si==0.3.6 si-server

Similar plugins