context

Plugin context from marketplace nonoun in kimgranlund/nonoun-plugins. Commands read from the repo's manifest — today.

Skills (12)

• classify-and-resolve

  Design and grow classify-and-resolve workflows: incoming items get a class from a closed set, with an explicit unclassified route, and go to a resolver with a declared contract, through a cascade of rules, observed evidence, decision models and humans, an autonomy matrix guarded by evaluation gates,

• context-fog-of-war

  Explain Codex, Claude Code, custom web-chat and runtime-harness bundle context-routing decisions with evidence-backed decision traces, routing diagnoses, and guidance improvements; statically audit a codebase or corpus's guidance contracts (catalogs, schemas, skills, prompts, checks, oracles); rende

• context-routing-review

  Review and test repository context routes for reusable-component authors and application consumers, including fresh-agent planning or implementation trials. Use after guidance reorganizations or demonstrated discovery gaps; not for ordinary UI implementation alone.

• controlled-natural-language

  Write and review high-stakes statements in a controlled natural language (CNL): a small, evidence-gated set of sentence patterns per document type, checked by a linter against the corpus vocabulary. Use for requirements, specs, ADRs, acceptance criteria, runbooks, SOPs, policies, compliance language

• decision-questions

  Design bounded questions for a decision model and act on the answers safely: question shape, domain-scoped options with authored contrast, threshold bands for choices and scores, compare mode with a margin for yes/no propositions, joint labels assembled in code, and a checker, suggester, voter role

• evaluation-system

  Decide whether something is good enough using evidence: case sets, graders, metrics and gates for prompts, agents, classifiers and resolvers, a skill's triggering, a linter's precision, generated artifacts or human review processes, grown in evidence-gated stages with a suite validator and a runner.

• generative-systems

  Design, use and improve generative systems grounded in authoritative kits, templates, contracts and assets. Use for context architecture, source fidelity, generation drift, reusable mechanics and bounded efficacy evaluation. Complements domain authoring skills; does not apply to every writing or cod

• icm-architect

  Design any process, idea, problem, or body of knowledge into an ICM (Interpretable Context Methodology) workspace — folder structure as agent architecture — or restructure an existing folder, repo, or vault into one. Use when the user wants to (1) turn a recurring workflow into an agent-runnable fol

• knowledge-modeling

  Define and evolve domain concepts, typed relationships, evidence-backed mappings and interpretation rules. Use to model knowledge, reconcile vocabularies, audit semantic decisions or preserve meaning across revisions; not for ordinary prose edits or documentation navigation alone.

• knowledge-organization-system

  Keep a corpus's vocabulary consistent with a lightweight KOS (knowledge organization system, a SKOS subset): one vocabulary.yaml of defined concepts, labels and links that grows only on evidence, plus a validator. Use for glossaries, terminology, naming conventions, "what do we call X", the same thi

• progressive-context

  Design, review, or refactor progressive-context systems for repositories, document corpora, knowledge bases and runtime harnesses. Use to improve how tasks discover sufficient authoritative context, with bounded traversal and evidence-backed route tests, without requiring a particular vendor or layo

• system-atlas

  Draws a whole system as large workflow diagrams from two directions and publishes them as one full-width artifact: an inventory of its machinery (skills, agents, commands, rules, hooks, checkers, evaluators, verifiers, ontologies, knowledge bases), a machinery map, journey marks, an enforcement matr

Similar plugins