Em 0025

Source content

{

"$schema": "https://epistemedia.com/schemas/task-contract-v1.json",

"id": "EM-0025",

"title": "Preserve the conceptual vocabulary and select the Case 002 editorial roadmap",

"status": "ready",

"change_class": "ordinary-implementation",

"objective": "Preserve the owner-directed Greek and modern epistemic vocabulary as durable product-design context, synthesize the supplied Case 002 strategy and pre-research into an explicit candidate ledger, select one bounded Case 002 research direction, and sequence a small library of later cases without admitting new empirical claims or beginning Case 002 research.",

"depends_on": [

"EM-0024"

],

"authority": {

"allowed_paths": [

"docs/**",

"runs/**"

],

"forbidden_paths": [

"constitution/**",

"policies/**",

"schemas/**",

"governance/events/**",

".github/**",

"research/**",

"catalog/**",

"src/**",

"tests/**",

"tasks/contracts/**",

"README.md"

]

},

"required_evaluation": [

"conceptual-vocabulary coverage and episteme-versus-truth boundary review",

"Case 002 candidate-ledger completeness against all supplied strategy and pre-research notes",

"explicit separation of editorial decision, pre-research leads, accepted evidence, and public claim admission",

"roadmap review for verdict-shape, audience, machinery, corpus-cost, licensing, and risk diversity",

"exact path-scope and prohibited-surface comparison",

"full deterministic make check with disclosure audit and unchanged generated state",

"independent review of the resulting documentation from a fresh clone"

],

"acceptance": [

"a durable conceptual-vocabulary document preserves the useful Greek and modern distinctions while stating that it is product orientation rather than an authoritative history of philosophy or a How We Know case",

"the vocabulary distinguishes truth, world state, belief, evidence, warrant, provenance, lineage, uncertainty, suspension, understanding, skill, judgment, and wisdom and connects them to concrete product implications",

"a standalone case-selection record synthesizes the supplied candidate arguments and pre-research without presenting those notes as accepted evidence",

"Case 002 is explicitly selected as a narrow agent-citation-lineage case whose empirical unit is captured agent traces rather than papers about agents",

"the selection fails closed if a reproducible public trace packet cannot support the claimed URL, span, derivation, and dependence counts",

"a sequenced library roadmap retains the strongest non-selected candidates and varies audience, verdict shape, and machinery rather than creating a one-topic debunking series",

"the current public feature, accepted Case 001 dossier, research packet, policies, compiler, workflows, and generated outputs remain unchanged",

"make check passes without changing accepted source state"

],

"limitations": [

"This task makes an editorial selection and preserves conceptual design context; it does not create, research, review, admit, feature, publish, or deploy Case 002.",

"Supplied model-generated pre-research is a lead ledger, not a source packet, citation, or evidentiary receipt; every empirical statement must be re-retrieved and verified under a later research contract.",

"The conceptual vocabulary does not establish a universal ontology of knowledge, settle historical interpretation, or make the word episteme a case or homepage hero.",

"No agent run, benchmark, medical claim, political claim, additional lens, public intake form, API/MCP activation, workflow mutation, DNS action, package publication, or release is authorized."

]

}