@citeark/agent 0.3.20

package/schemas/research.schema.json

@@ -0,0 +1,1044 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://citeark.com/schemas/research/structure-1.0.json",
+  "title": "CiteArk structured research",
+  "type": "object",
+  "required": ["schemaVersion", "work", "sources", "claims", "experiments", "lineage", "coverage", "license", "provenance"],
+  "properties": {
+    "reading": {
+      "type": "object",
+      "required": [
+        "schemaVersion",
+        "overview",
+        "mainline",
+        "branches"
+      ],
+      "properties": {
+        "schemaVersion": {
+          "const": "1.0"
+        },
+        "overview": {
+          "type": "string",
+          "minLength": 1
+        },
+        "mainline": {
+          "type": "array",
+          "minItems": 1,
+          "uniqueItems": true,
+          "items": {
+            "type": "string",
+            "pattern": "^(object|claim|hypothesis|source):.+"
+          }
+        },
+        "branches": {
+          "type": "array",
+          "items": {
+            "type": "object",
+            "required": [
+              "parent",
+              "children"
+            ],
+            "properties": {
+              "parent": {
+                "type": "string",
+                "pattern": "^(object|claim|hypothesis|source):.+"
+              },
+              "children": {
+                "type": "array",
+                "uniqueItems": true,
+                "items": {
+                  "type": "string",
+                  "pattern": "^(object|claim|hypothesis|source):.+"
+                }
+              }
+            }
+          }
+        },
+        "labels": {
+          "type": "array",
+          "items": {
+            "type": "object",
+            "required": [
+              "target",
+              "title"
+            ],
+            "properties": {
+              "target": {
+                "type": "string",
+                "pattern": "^(object|claim|hypothesis|source):.+"
+              },
+              "title": {
+                "type": "string",
+                "minLength": 1
+              },
+              "subtype": {
+                "type": "string"
+              }
+            }
+          }
+        }
+      }
+    },
+    "researchRelations": {
+      "type": "array",
+      "items": {
+        "type": "object",
+        "required": [
+          "id",
+          "from",
+          "to",
+          "predicate",
+          "basis",
+          "sourceLocators"
+        ],
+        "properties": {
+          "id": {
+            "type": "string",
+            "minLength": 1
+          },
+          "from": {
+            "type": "string",
+            "pattern": "^(object|claim|hypothesis|source):.+"
+          },
+          "to": {
+            "type": "string",
+            "pattern": "^(object|claim|hypothesis|source):.+"
+          },
+          "predicate": {
+            "enum": [
+              "about",
+              "describes",
+              "derivedFrom",
+              "assertedIn",
+              "attributedTo",
+              "plannedFor",
+              "requires",
+              "used",
+              "generated",
+              "follows",
+              "associatedWith",
+              "partOf",
+              "evidenceFor",
+              "assesses",
+              "basedOn"
+            ]
+          },
+          "basis": {
+            "enum": [
+              "declared",
+              "inferred"
+            ]
+          },
+          "sourceLocators": {
+            "type": "array",
+            "minItems": 1,
+            "items": {
+              "type": "object",
+              "required": [
+                "sourceId",
+                "locator"
+              ],
+              "properties": {
+                "sourceId": {
+                  "type": "string"
+                },
+                "locator": {
+                  "type": "string",
+                  "minLength": 1
+                }
+              }
+            }
+          },
+          "note": {
+            "type": "string"
+          }
+        }
+      }
+    },
+    "researchObjects": {
+      "type": "array",
+      "description": "Independently addressable scientific objects, reused by explicit identity across routes.",
+      "items": {
+        "type": "object",
+        "required": [
+          "id",
+          "role",
+          "title"
+        ],
+        "properties": {
+          "id": {
+            "type": "string",
+            "minLength": 1
+          },
+          "sourceId": {
+            "type": "string",
+            "description": "Identity of an existing material source. Use sourceLocators for mentions; a derived split gets its own identity."
+          },
+          "role": {
+            "enum": [
+              "dataset",
+              "sample",
+              "model",
+              "checkpoint",
+              "code",
+              "method",
+              "observation",
+              "material", "question", "concept", "premise", "argument", "research-plan", "research-activity", "source-assessment", "person", "organization", "instrument"
+            ]
+          },
+          "title": {
+            "type": "string",
+            "minLength": 1
+          },
+          "description": {
+            "type": "string"
+          },
+          "observationKind": {"enum": ["scalar", "table", "matrix", "series", "collection"], "description": "Source-grounded representation of an observation collection; not its scientific conclusion."},
+          "basis": {
+            "enum": [
+              "declared",
+              "inferred"
+            ]
+          },
+          "identity": {
+            "type": "object",
+            "description": "Only source-grounded identities: URI, revision, digest, split, architecture, seed. Omit unknown values."
+          },
+          "conditions": {},
+          "sourceLocators": {
+            "type": "array",
+            "items": {
+              "type": "object",
+              "required": [
+                "sourceId",
+                "locator"
+              ],
+              "properties": {
+                "sourceId": {
+                  "type": "string"
+                },
+                "locator": {
+                  "type": "string",
+                  "minLength": 1
+                }
+              }
+            }
+          },
+          "derivedFromIds": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {
+              "type": "string",
+              "minLength": 1
+            }
+          },
+          "prospective": {
+            "type": "boolean",
+            "description": "True for expected objects that this planned run has not produced."
+          },
+          "limitations": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            }
+          }
+        }
+      }
+    },
+    "compilationStage": { "const": "inventory", "description": "A source inventory awaiting reproduction preparation; no executable experiments or feasibility verdicts." },
+    "sourceInventory": {
+      "type": "object", "required": ["digest", "research"],
+      "properties": {
+        "digest": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
+        "research": { "$ref": "#", "description": "The unchanged initial source reading, retained even when later interpretation is corrected." },
+        "revisions": { "type": "array", "items": { "type": "object", "required": ["claimId", "claim", "reason", "sources"],
+          "properties": { "claimId": { "type": "string" }, "claim": { "$ref": "#/properties/claims/items" },
+            "reason": { "type": "string", "minLength": 1 }, "sources": { "type": "array", "minItems": 1,
+              "items": { "type": "object", "required": ["sourceId", "locator"], "properties": {
+                "sourceId": { "type": "string" }, "locator": { "type": "string", "minLength": 1 }
+              } } } }
+        } }
+      }
+    },
+    "reproductionScope": { "enum": ["low", "medium", "high"], "description": "Requested initial-selection preset; not scientific support, admission, fidelity or budget." },
+    "experimentSelection": {
+      "description": "Agent-authored catalog groups and recommendations for initial experiment checkboxes. Groups organize the review UI only; neither groups nor recommendations remove experiments or constrain later user selection.",
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["presets"],
+      "properties": {
+        "importance": {
+          "type": "array",
+          "description": "Scientific importance independent of cost, feasibility, confidence or evidence verdict. Optional presentation metadata; absent means unrated.",
+          "items": {
+            "type": "object",
+            "additionalProperties": false,
+            "required": [
+              "experimentId",
+              "stars",
+              "rationale"
+            ],
+            "properties": {
+              "experimentId": {
+                "type": "string"
+              },
+              "stars": {
+                "type": "integer",
+                "minimum": 1,
+                "maximum": 3
+              },
+              "rationale": {
+                "type": "string",
+                "minLength": 1
+              },
+              "headline": {
+                "type": "string",
+                "minLength": 1,
+                "maxLength": 180
+              }
+            }
+          }
+        },
+        "groups": {
+          "type": "array",
+          "items": { "$ref": "#/$defs/selectionGroup" }
+        },
+        "presets": {
+          "type": "object",
+          "additionalProperties": false,
+          "properties": {
+            "low": { "$ref": "#/$defs/selectionPreset" },
+            "medium": { "$ref": "#/$defs/selectionPreset" },
+            "high": { "$ref": "#/$defs/selectionPreset" }
+          }
+        }
+      }
+    },
+    "schemaVersion": { "const": "1.0" },
+    "work": {
+      "type": "object",
+      "required": ["id", "title", "abstract", "significance"],
+      "properties": {
+        "id": { "type": "string", "minLength": 1 },
+        "title": { "type": "string", "minLength": 1 },
+        "abstract": {
+          "description": "Objective account of the paper, rewritten in the compiler's own words (never copied from the paper's original abstract); 80–120 words covering the central problem, method, and headline result.",
+          "type": "string",
+          "minLength": 1,
+          "maxLength": 1200
+        },
+        "significance": {
+          "description": "Plain-language AI interpretation of what the paper changes, who may benefit, and the main boundary of that value. Editorial context only; never reproduction evidence or an Assessment.",
+          "type": "string",
+          "minLength": 1,
+          "maxLength": 4000
+        },
+        "authors": {
+          "description": "Ordered paper author names extracted from the fixed paper snapshot. Newly compiled research must include this field; use an empty array only when the byline is genuinely unreadable.",
+          "type": "array",
+          "maxItems": 500,
+          "items": { "type": "string", "minLength": 1, "maxLength": 500 }
+        },
+        "subjects": {
+          "description": "CiteArk controlled subject tags: the full arXiv category taxonomy, slug = category id lowercased with '.' replaced by '-' (cs.LG → cs-lg). Mirror of src/domain/subject-tags.ts.",
+          "type": "array",
+          "items": {
+            "enum": [
+              "cs-ai", "cs-ar", "cs-cc", "cs-ce", "cs-cg", "cs-cl", "cs-cr", "cs-cv", "cs-cy", "cs-db",
+              "cs-dc", "cs-dl", "cs-dm", "cs-ds", "cs-et", "cs-fl", "cs-gl", "cs-gr", "cs-gt", "cs-hc",
+              "cs-ir", "cs-it", "cs-lg", "cs-lo", "cs-ma", "cs-mm", "cs-ms", "cs-na", "cs-ne", "cs-ni",
+              "cs-oh", "cs-os", "cs-pf", "cs-pl", "cs-ro", "cs-sc", "cs-sd", "cs-se", "cs-si", "cs-sy",
+              "econ-em", "econ-gn", "econ-th",
+              "eess-as", "eess-iv", "eess-sp", "eess-sy",
+              "math-ac", "math-ag", "math-ap", "math-at", "math-ca", "math-co", "math-ct", "math-cv",
+              "math-dg", "math-ds", "math-fa", "math-gm", "math-gn", "math-gr", "math-gt", "math-ho",
+              "math-it", "math-kt", "math-lo", "math-mg", "math-mp", "math-na", "math-nt", "math-oa",
+              "math-oc", "math-pr", "math-qa", "math-ra", "math-rt", "math-sg", "math-sp", "math-st",
+              "astro-ph-co", "astro-ph-ep", "astro-ph-ga", "astro-ph-he", "astro-ph-im", "astro-ph-sr",
+              "cond-mat-dis-nn", "cond-mat-mes-hall", "cond-mat-mtrl-sci", "cond-mat-other",
+              "cond-mat-quant-gas", "cond-mat-soft", "cond-mat-stat-mech", "cond-mat-str-el", "cond-mat-supr-con",
+              "gr-qc", "hep-ex", "hep-lat", "hep-ph", "hep-th", "math-ph",
+              "nlin-ao", "nlin-cd", "nlin-cg", "nlin-ps", "nlin-si",
+              "nucl-ex", "nucl-th",
+              "physics-acc-ph", "physics-ao-ph", "physics-app-ph", "physics-atm-clus", "physics-atom-ph",
+              "physics-bio-ph", "physics-chem-ph", "physics-class-ph", "physics-comp-ph", "physics-data-an",
+              "physics-ed-ph", "physics-flu-dyn", "physics-gen-ph", "physics-geo-ph", "physics-hist-ph",
+              "physics-ins-det", "physics-med-ph", "physics-optics", "physics-plasm-ph", "physics-pop-ph",
+              "physics-soc-ph", "physics-space-ph",
+              "quant-ph",
+              "q-bio-bm", "q-bio-cb", "q-bio-gn", "q-bio-mn", "q-bio-nc", "q-bio-ot", "q-bio-pe",
+              "q-bio-qm", "q-bio-sc", "q-bio-to",
+              "q-fin-cp", "q-fin-ec", "q-fin-gn", "q-fin-mf", "q-fin-pm", "q-fin-pr", "q-fin-rm",
+              "q-fin-st", "q-fin-tr",
+              "stat-ap", "stat-co", "stat-me", "stat-ml", "stat-ot", "stat-th"
+            ]
+          }
+        }
+      }
+    },
+    "sources": {
+      "type": "array",
+      "minItems": 2,
+      "items": {
+        "type": "object",
+        "required": ["id", "kind", "uri"],
+        "properties": {
+          "id": { "type": "string", "minLength": 1 },
+          "kind": { "enum": ["paper", "repository", "dataset", "supplement", "checkpoint"] },
+          "uri": { "type": "string", "minLength": 1 },
+          "version": { "type": "string" },
+          "digest": { "type": "string" }
+        }
+      }
+    },
+    "hypotheses": {
+      "type": "array",
+      "description": "Source-proposed explanations, distinct from empirical claims and scientific assessments. No execution disposition or support verdict.",
+      "items": {
+        "type": "object",
+        "required": [
+          "id",
+          "statement",
+          "sourceLocator"
+        ],
+        "properties": {
+          "id": {
+            "type": "string",
+            "minLength": 1
+          },
+          "statement": {
+            "type": "string",
+            "minLength": 1
+          },
+          "sourceLocator": {
+            "type": "object",
+            "required": [
+              "sourceId",
+              "locator"
+            ],
+            "properties": {
+              "sourceId": {
+                "type": "string"
+              },
+              "locator": {
+                "type": "string",
+                "minLength": 1
+              }
+            }
+          },
+          "objectIds": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {
+              "type": "string"
+            }
+          },
+          "conditions": {},
+          "limitations": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            }
+          }
+        }
+      }
+    },
+    "claims": {
+      "type": "array",
+      "minItems": 1,
+      "items": {
+        "type": "object",
+        "required": ["id", "statement", "type", "sourceLocator", "reportedMeasurements", "reproduction"],
+        "properties": {
+          "objectIds": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {
+              "type": "string",
+              "minLength": 1
+            }
+          },
+          "id": { "type": "string", "minLength": 1 },
+          "versionId": { "type": "string" },
+          "statement": { "type": "string", "minLength": 1 },
+          "type": { "enum": ["finding", "method", "measurement", "limitation"] },
+          "icon": { "type": "string", "description": "Optional presentation icon from the CiteArk claim icon whitelist (kebab-case lucide name) matching the claim's subject matter, e.g. clock for training-time claims or gauge for metric claims. Decorative only; not part of evidence semantics." },
+          "sourceLocator": { "$ref": "#/$defs/locator" },
+          "reportedMeasurements": {
+            "type": "array",
+            "items": {
+              "allOf": [
+                { "$ref": "#/$defs/measurement" },
+                { "type": "object", "required": ["id", "value", "sourceLocator"], "properties": { "observationId": {"type": "string", "minLength": 1, "description": "Explicit source observation collection identity in researchObjects; numerical rows retain their own measurement IDs."}, "id": { "type": "string", "minLength": 1 }, "value": { "type": "number" }, "verificationTolerance": { "description": "Pre-execution absolute tolerance in the measurement unit. Required when no safe unit default exists and a strict comparison is intended.", "type": "number", "minimum": 0 }, "evidenceBasis": { "description": "How the paper-side numeric target was obtained.", "enum": ["reported_numeric", "digitized_figure", "derived_reported_relation"] }, "extractionUncertainty": { "description": "Absolute paper-side extraction uncertainty in the measurement unit, especially for values digitized from figures.", "type": "number", "minimum": 0 }, "sourceLocator": { "$ref": "#/$defs/locator" }, "dimensions": { "type": "object" } } }
+              ]
+            }
+          },
+          "reproduction": {
+            "type": "object",
+            "required": ["status"],
+            "properties": {
+              "status": { "description": "deferred means no experiment is scheduled and feasibility is unproven; blocked requires a concrete blocker with kind and source-backed evidence, never merely a missing experiment; planned requires at least one directly relevant automatically runnable measurement and may disclose partial coverage in reason; blocked is used only when no meaningful paper-bound experiment can run because a required scientific input, safe decision rule, access right, or feasible execution profile is unavailable; not_applicable is reserved for actual limitations, non-empirical method context, or claims derived from another represented claim.", "enum": ["planned", "deferred", "blocked", "not_applicable"] },
+              "scopeExclusions": { "type": "array", "items": { "type": "object", "required": ["measurementIds", "minimumScope", "reason"], "properties": { "measurementIds": { "type": "array", "minItems": 1, "uniqueItems": true, "items": { "type": "string" } }, "minimumScope": { "enum": ["medium", "high"] }, "reason": { "type": "string", "minLength": 1 } } } },
+              "blocker": { "type": "object", "required": ["kind", "evidence"], "properties": { "kind": { "enum": ["scientific_input", "access", "resources", "budget", "decision_rule"] }, "evidence": { "type": "string", "minLength": 1 }, "measurementIds": { "type": "array", "items": { "type": "string", "minLength": 1 }, "uniqueItems": true } } },
+              "implementationOrigin": { "enum": ["official", "citeark_reconstruction", "insufficient"] },
+              "reason": { "type": "string", "minLength": 1 },
+              "derivedFromClaimIds": { "type": "array", "items": { "type": "string", "minLength": 1 } }
+            }
+          }
+        }
+      }
+    },
+    "experiments": {
+      "type": "array",
+      "items": {
+        "type": "object",
+        "required": ["id", "title", "claimIds", "reproductionLevel", "repository", "protocol", "measurements"],
+        "anyOf": [
+          {"properties": {"measurements": {"minItems": 1}}},
+          {"required": ["observationTargets"], "properties": {"observationTargets": {"minItems": 1}}}
+        ],
+        "properties": {
+          "researchIntent": {
+            "type": "object",
+            "required": [
+              "question",
+              "comparison",
+              "criterion",
+              "from",
+              "to"
+            ],
+            "properties": {
+              "question": {
+                "type": "string",
+                "minLength": 1
+              },
+              "comparison": {
+                "type": "string",
+                "minLength": 1
+              },
+              "criterion": {
+                "type": "string",
+                "minLength": 1
+              },
+              "from": {
+                "type": "string",
+                "pattern": "^(object|claim|hypothesis|source):.+"
+              },
+              "to": {
+                "type": "string",
+                "pattern": "^(object|claim|hypothesis|source):.+"
+              },
+              "context": {
+                "type": "array",
+                "items": {
+                  "type": "string",
+                  "pattern": "^(object|claim|hypothesis|source):.+"
+                }
+              }
+            }
+          },
+          "steps": {
+            "type": "array",
+            "description": "Scientific method stages, not shell commands or Cartesian-product execution instances. Complete seed/checkpoint/layer axes belong in the stage description and expected collection conditions. Shared IDs require exactly the same inputs, outputs, recipe and workload identities.",
+            "items": {
+              "type": "object",
+              "required": [
+                "id",
+                "title",
+                "kind",
+                "inputIds",
+                "outputIds"
+              ],
+              "properties": {
+                "id": {
+                  "type": "string",
+                  "minLength": 1
+                },
+                "title": {
+                  "type": "string",
+                  "minLength": 1
+                },
+                "kind": {
+                  "enum": [
+                    "preparation",
+                    "collection",
+                    "preprocessing",
+                    "training",
+                    "evaluation",
+                    "analysis"
+                  ]
+                },
+                "description": {
+                  "type": "string"
+                },
+                "inputIds": {
+                  "type": "array",
+                  "uniqueItems": true,
+                  "items": {
+                    "type": "string",
+                    "minLength": 1
+                  }
+                },
+                "outputIds": {
+                  "type": "array",
+                  "uniqueItems": true,
+                  "items": {
+                    "type": "string",
+                    "minLength": 1
+                  }
+                },
+                "dependsOn": {
+                  "type": "array",
+                  "uniqueItems": true,
+                  "items": {
+                    "type": "string",
+                    "minLength": 1
+                  }
+                },
+                "claimIds": {
+                  "type": "array",
+                  "uniqueItems": true,
+                  "items": {
+                    "type": "string",
+                    "minLength": 1
+                  }
+                },
+                "workloadIds": {
+                  "type": "array",
+                  "uniqueItems": true,
+                  "items": {
+                    "type": "string",
+                    "minLength": 1
+                  }
+                }
+              }
+            }
+          },
+          "inputIds": {
+            "type": "array",
+            "uniqueItems": true,
+            "items": {
+              "type": "string",
+              "minLength": 1
+            }
+          },
+          "id": { "type": "string", "minLength": 1 },
+          "versionId": { "type": "string" },
+          "title": { "type": "string", "minLength": 1 },
+          "claimIds": {
+            "description": "Claims tested by this experiment. An experiment may bind only the reported measurements it actually executes; multiple experiments may collectively cover one claim.",
+            "type": "array",
+            "minItems": 1,
+            "items": { "type": "string" }
+          },
+          "reproductionScope": { "enum": ["low", "medium", "high"] },
+          "requiredReproductionScope": { "enum": ["low", "medium", "high"], "description": "Minimum scope required by the actual work, across all disciplines." },
+          "reproductionLevel": { "enum": ["directional", "official-checkpoint", "full-training", "artifact-evaluation", "experiment-rerun", "workflow-rebuild"] },
+          "reconstructionFidelity": {
+            "description": "Scientific protocol fidelity, independent of whether the implementation is official or a CiteArk reconstruction.",
+            "enum": ["exact", "faithful", "approximate", "proxy"]
+          },
+          "implementationOrigin": { "enum": ["official", "citeark_reconstruction"] },
+          "repository": {
+            "type": "object",
+            "required": ["url", "commit"],
+            "properties": {
+              "implementationOrigin": { "enum": ["official", "citeark_reconstruction"] },
+              "url": { "type": "string", "minLength": 1 },
+              "commit": { "type": "string", "minLength": 1 }
+            }
+          },
+          "protocol": {
+            "type": "object",
+            "required": ["entrypoint", "requiredCommandFragments"],
+            "properties": {
+              "workload": {
+                "description": "Complete scientific execution units, grouped by materially different cost or method; probes must execute full units on the allocated device.",
+                "type": "object",
+                "additionalProperties": false,
+                "required": ["unit", "groups"],
+                "properties": {
+                  "unit": { "type": "string", "minLength": 1 },
+                  "groups": {
+                    "type": "array", "minItems": 1,
+                    "items": {
+                      "type": "object", "additionalProperties": false,
+                      "required": ["id", "units", "description"],
+                      "properties": {
+                        "id": { "type": "string", "minLength": 1 },
+                        "units": { "type": "integer", "minimum": 1 },
+                        "description": { "type": "string", "minLength": 1 }
+                      }
+                    }
+                  }
+                }
+              },
+
+              "executionMode": {
+                "description": "agent_planned lets the execution Agent choose and revise concrete commands while preserving the immutable scientific objective and measurement scope. fixed_entrypoint preserves legacy exact-command behavior.",
+                "enum": ["agent_planned", "fixed_entrypoint"]
+              },
+              "objective": { "type": "string", "minLength": 1 },
+              "entrypoint": { "type": "string", "minLength": 1 },
+              "requiredCommandFragments": { "type": "array", "items": { "type": "string", "minLength": 1 } },
+              "instructions": { "type": "string" },
+              "benchmark": {
+                "description": "Complete prospective protocol for hardware-sensitive timing, throughput, memory, energy, or speedup measurements.",
+                "type": "object",
+                "additionalProperties": false,
+                "required": ["batchSize", "dtype", "warmupIterations", "measurementIterations", "timer", "synchronization", "inputConstruction", "implementations"],
+                "properties": {
+                  "device": { "enum": ["cpu", "gpu"], "description": "Device required by the measurement itself. A GPU benchmark never falls back to CPU." },
+                  "referenceHardware": { "type": "string", "description": "Paper hardware or explicit independent benchmark reference; another GPU does not establish equivalence." },
+                  "hardwareBinding": { "enum": ["required", "not_required", "unknown"], "description": "Whether matching the paper's hardware and scale is scientifically required to assess the bound reported measurement. unknown is not automatically executable." },
+                  "hardwareBindingRationale": { "type": "string", "minLength": 1, "description": "Source-grounded explanation of why the reported measurement is or is not bound to the paper hardware and scale." },
+                  "hardwareBindingSources": { "type": "array", "minItems": 1, "items": { "$ref": "#/$defs/locator" }, "description": "Paper or repository locations that establish the hardware-binding decision." },
+                  "comparisonIntent": { "enum": ["reference_conditions", "cross_hardware"], "description": "Intended scientific comparison, not a comparability verdict. Specify whether reference conditions or a limited cross-hardware question are being tested." },
+                  "comparisonRationale": { "type": "string", "minLength": 1, "description": "Source-grounded explanation of what the allocated hardware can establish and which original performance conditions remain untested." },
+                  "requiredGpuTypes": { "type": "array", "items": { "type": "string", "minLength": 1 }, "description": "Exact catalog GPU type constraints only when scientifically required; never encode merely preferred types here." },
+                  "requiredGpuCount": { "type": "integer", "minimum": 1, "description": "Paper-bound GPU count required for the reported measurement, distinct from a cheaper exploratory allocation." },
+                  "requiredVramGb": { "type": "number", "exclusiveMinimum": 0, "description": "Paper-bound minimum VRAM in GiB per GPU when the reported measurement requires it." },
+                  "batchSize": {
+                    "oneOf": [
+                      { "type": "integer", "minimum": 1 },
+                      { "type": "array", "minItems": 1, "uniqueItems": true, "items": { "type": "integer", "minimum": 1 } }
+                    ]
+                  },
+                  "dtype": { "type": "string", "minLength": 1 },
+                  "warmupIterations": { "type": "integer", "minimum": 1 },
+                  "measurementIterations": { "type": "integer", "minimum": 1 },
+                  "timer": { "type": "string", "minLength": 1 },
+                  "synchronization": { "type": "string", "minLength": 1 },
+                  "inputConstruction": { "type": "string", "minLength": 1 },
+                  "implementations": {
+                    "type": "array",
+                    "minItems": 1,
+                    "items": {
+                      "type": "object",
+                      "additionalProperties": false,
+                      "required": ["id", "callable", "provenance"],
+                      "properties": {
+                        "id": { "type": "string", "minLength": 1 },
+                        "callable": { "type": "string", "minLength": 1 },
+                        "provenance": { "type": "string", "minLength": 1 }
+                      }
+                    }
+                  }
+                }
+              },
+              "preflight": {
+                "description": "Bounded operational gate that must pass before the expensive scientific command begins. Optional only for legacy Research Plan 1.0 records.",
+                "type": "object",
+                "additionalProperties": false,
+                "required": ["schemaVersion", "source", "maxMinutes", "evidencePath", "checks", "assets"],
+                "properties": {
+                  "schemaVersion": { "enum": ["0.1", "0.2", "0.3"] },
+                  "source": { "enum": ["declared", "defaulted"] },
+                  "maxMinutes": { "type": "number", "minimum": 1, "maximum": 240 },
+                  "evidencePath": {
+                    "type": "string",
+                    "minLength": 1,
+                    "pattern": "^(?!/)(?!\\.\\.(?:/|$))(?!.*\\/\\.\\.(?:/|$))[^\\\\]+$"
+                  },
+                  "checks": {
+                    "type": "array",
+                    "minItems": 1,
+                    "items": {
+                      "type": "object",
+                      "additionalProperties": false,
+                      "required": ["id", "kind", "successCriterion"],
+                      "properties": {
+                        "id": { "type": "string", "minLength": 1 },
+                        "kind": { "enum": ["runtime", "dependencies", "dataset_loader", "assets", "entrypoint", "parser", "smoke_test", "duration"] },
+                        "command": { "type": "string", "minLength": 1 },
+                        "successCriterion": { "type": "string", "minLength": 1 }
+                      }
+                    }
+                  },
+                  "assets": {
+                    "type": "array",
+                    "items": {
+                      "type": "object",
+                      "additionalProperties": false,
+                      "required": ["id", "kind", "identity", "availability", "required", "verification"],
+                      "properties": {
+                        "id": { "type": "string", "minLength": 1 },
+                        "kind": { "enum": ["dataset", "checkpoint", "model", "repository", "other"] },
+                        "identity": { "type": "string", "minLength": 1 },
+                        "availability": { "enum": ["bundled", "public_verified", "public_unverified", "private", "unknown"] },
+                        "required": { "type": "boolean" },
+                        "verification": { "type": "string", "minLength": 1 },
+                        "scientificRole": { "enum": ["execution_required", "identity_only", "provenance_only", "optional"] },
+                        "acquisition": { "$ref": "#/$defs/assetAcquisition" },
+                        "sourceEvidence": {
+                          "type": "array",
+                          "items": { "$ref": "#/$defs/assetSourceEvidence" }
+                        },
+                        "sourceUrls": {
+                          "type": "array",
+                          "minItems": 1,
+                          "uniqueItems": true,
+                          "items": { "type": "string", "format": "uri", "pattern": "^https?://" }
+                        },
+                        "registryDatasetId": { "type": "string", "minLength": 1 }
+                      }
+                    }
+                  },
+                  "requirements": {
+                    "description": "Complete eight-row scientific material inventory. Required for newly compiled preflight 0.3 plans.",
+                    "type": "array",
+                    "minItems": 8,
+                    "maxItems": 8,
+                    "items": { "$ref": "#/$defs/assetRequirement" }
+                  }
+                }
+              }
+            }
+          },
+          "observationTargets": {
+            "description": "Proposed comparisons of declared source observation collections. They preserve methods in the planning catalog but do not create numerical verification targets or automatic execution admission. Use only when a deterministic comparison is unresolved.",
+            "type": "array",
+            "items": {
+              "type": "object", "additionalProperties": false,
+              "required": ["claimId", "observationId", "comparison", "limitation"],
+              "properties": {
+                "claimId": {"type": "string", "minLength": 1},
+                "observationId": {"type": "string", "minLength": 1},
+                "comparison": {"type": "string", "minLength": 1},
+                "limitation": {"type": "object", "additionalProperties": false, "required": ["kind", "evidence"],
+                  "properties": {"kind": {"const": "decision_rule"}, "evidence": {"type": "string", "minLength": 1}}}
+              }
+            }
+          },
+          "measurements": {
+            "description": "Numerical targets only. Every reportedMeasurementId must resolve to an actual claims[].reportedMeasurements[].id, never an observation object ID. May be empty for a proposed observation-only method.",
+            "type": "array",
+            "items": {
+              "allOf": [
+                { "$ref": "#/$defs/measurement" },
+                {
+                  "type": "object",
+                  "required": ["reportedMeasurementId", "parser"],
+                  "properties": {
+                    "reportedMeasurementId": { "type": "string", "minLength": 1 },
+                    "parser": {
+                      "type": "object",
+                      "required": ["id", "version", "evidencePath"],
+                      "properties": {
+                        "id": { "type": "string", "minLength": 1 },
+                        "version": { "type": "string", "minLength": 1 },
+                        "evidencePath": {
+                          "type": "string",
+                          "minLength": 1,
+                          "pattern": "^(?!/)(?!\\.\\.(?:/|$))(?!.*\\/\\.\\.(?:/|$))[^\\\\]+$"
+                        },
+                        "config": { "type": "object" }
+                      }
+                    }
+                  }
+                }
+              ]
+            }
+          },
+          "agent": { "type": "object" },
+          "compute": { "$ref": "#/$defs/computeRequirement" },
+          "environment": { "type": "object", "properties": {"timeoutMinutes": {"type": "number", "exclusiveMinimum": 0, "description": "Finite execution timeout, separate from the complete workload estimate; bounded by computeContext.executionRuntime.timeoutCapsMinutes when supplied."}} }
+        }
+      }
+    },
+    "lineage": {
+      "type": "array",
+      "items": {
+        "type": "object",
+        "required": ["from", "to", "relation"],
+        "properties": {
+          "from": { "type": "string" },
+          "to": { "type": "string" },
+          "relation": { "type": "string" }
+        }
+      }
+    },
+    "coverage": {
+      "type": "object",
+      "required": ["items"],
+      "properties": {
+        "items": {
+          "type": "array",
+          "minItems": 1,
+          "items": {
+            "type": "object",
+            "required": ["id", "description", "sourceLocator", "status", "claimIds"],
+            "properties": {
+              "id": { "type": "string", "minLength": 1 },
+              "description": { "type": "string", "minLength": 1 },
+              "sourceLocator": { "$ref": "#/$defs/locator" },
+              "status": { "enum": ["represented", "deferred", "blocked", "not_applicable"] },
+              "claimIds": { "type": "array", "items": { "type": "string", "minLength": 1 } },
+              "reason": { "type": "string", "minLength": 1 }
+            }
+          }
+        }
+      }
+    },
+    "license": {
+      "type": "object",
+      "required": ["paper", "code", "executionAllowed"],
+      "properties": {
+        "paper": { "type": "string" },
+        "code": { "type": "string" },
+        "dataset": { "type": "string" },
+        "checkpoint": { "type": "string" },
+        "executionAllowed": { "type": "boolean" },
+        "redistribution": { "type": "string" }
+      }
+    },
+    "provenance": {
+      "type": "object",
+      "required": ["compiler", "compiledAt", "reviewStatus"],
+      "properties": {
+        "compiler": { "type": "string" },
+        "compiledAt": { "type": "string" },
+        "reviewStatus": { "enum": ["agent-draft", "human-reviewed"] },
+        "inputs": { "type": "array", "items": { "type": "string" } },
+        "normalizationWarnings": {
+          "type": "array",
+          "description": "Deterministic compiler-boundary repairs or conservative degradations applied by CiteArk.",
+          "items": { "type": "string" }
+        }
+      }
+    }
+  },
+  "$defs": {
+    "selectionGroup": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["id", "title", "description", "experimentIds"],
+      "properties": {
+        "id": { "type": "string", "minLength": 1 },
+        "title": { "type": "string", "minLength": 1 },
+        "description": { "type": "string", "minLength": 1 },
+        "experimentIds": { "type": "array", "minItems": 1, "uniqueItems": true, "items": { "type": "string", "minLength": 1 } }
+      }
+    },
+    "selectionPreset": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["experimentIds", "rationale"],
+      "properties": {
+        "experimentIds": { "type": "array", "uniqueItems": true, "items": { "type": "string", "minLength": 1 } },
+        "rationale": { "type": "string", "minLength": 1 }
+      }
+    },
+    "assetRequirement": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["kind", "required", "reason", "assetIds", "checkIds", "repository"],
+      "properties": {
+        "kind": { "enum": ["author_code", "model", "tokenizer", "training_data", "evaluation_data", "evaluator", "configuration", "dependencies"] },
+        "required": { "type": "boolean" },
+        "reason": { "type": "string", "minLength": 1 },
+        "assetIds": { "type": "array", "uniqueItems": true, "items": { "type": "string", "minLength": 1 } },
+        "checkIds": { "type": "array", "uniqueItems": true, "items": { "type": "string", "minLength": 1 } },
+        "repository": { "type": "boolean" }
+      }
+    },
+    "assetAcquisition": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["mode"],
+      "properties": {
+        "mode": { "enum": ["auto", "metadata", "selective", "full", "stream"] },
+        "revision": { "type": "string", "minLength": 1 },
+        "requiredPaths": { "type": "array", "description": "Backward-compatible selective prefetch hints. Entries may describe alternative formats; an unmatched entry does not block the reproduction Agent.", "uniqueItems": true, "items": { "type": "string", "minLength": 1 } },
+        "excludedPaths": { "type": "array", "uniqueItems": true, "items": { "type": "string", "minLength": 1 } },
+        "expectedDownloadBytes": { "type": "integer", "minimum": 0 },
+        "expectedExpandedBytes": { "type": "integer", "minimum": 0 },
+        "maximumDownloadBytes": { "type": "integer", "minimum": 1 },
+        "maximumExpandedBytes": { "type": "integer", "minimum": 1 },
+        "supportsStreaming": { "type": "boolean" }
+      }
+    },
+    "assetSourceEvidence": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["source", "locator"],
+      "properties": {
+        "source": { "enum": ["paper", "repository", "dataset_registry"] },
+        "locator": { "type": "string", "minLength": 1 },
+        "path": { "type": "string", "minLength": 1 },
+        "commit": { "type": "string", "minLength": 1 },
+        "resolver": { "enum": ["exact_url", "append_path", "huggingface_model", "huggingface_dataset"] },
+        "suffix": { "type": "string", "minLength": 1 }
+      }
+    },
+    "locator": {
+      "type": "object",
+      "required": ["sourceId", "locator"],
+      "properties": {
+        "sourceId": { "type": "string", "minLength": 1 },
+        "locator": { "type": "string", "minLength": 1 }
+      }
+    },
+    "measurement": {
+      "type": "object",
+      "required": ["metric", "unit"],
+      "properties": {
+        "objectIds": {
+          "type": "array",
+          "uniqueItems": true,
+          "items": {
+            "type": "string",
+            "minLength": 1
+          }
+        },
+        "metric": { "type": "string", "minLength": 1 },
+        "quantityKind": { "enum": ["device_performance", "operation_count", "numerical_result"], "description": "Scientific quantity semantics used for hardware admission; metric spelling alone is not authoritative." },
+        "unit": { "type": "string", "minLength": 1 },
+        "value": { "type": "number" },
+        "aggregation": { "type": "string" },
+        "evidenceBasis": { "enum": ["reported_numeric", "digitized_figure", "derived_reported_relation"] },
+        "extractionUncertainty": { "type": "number", "minimum": 0 }
+      }
+    },
+    "computeRequirement": {
+      "type": "object",
+      "additionalProperties": false,
+      "required": ["accelerator", "cpuCores", "memoryGb", "gpuCount", "cpuFallbackAllowed", "rationale"],
+      "allOf": [
+        {
+          "if": { "properties": { "accelerator": { "enum": ["optional", "required"] } } },
+          "then": { "properties": { "gpuCount": { "exclusiveMinimum": 0 } } }
+        }
+      ],
+      "properties": {
+        "accelerator": { "enum": ["none", "optional", "required"] },
+        "cpuCores": { "type": "number", "exclusiveMinimum": 0 },
+        "memoryGb": { "type": "number", "exclusiveMinimum": 0 },
+        "gpuCount": { "type": "number", "minimum": 0 },
+        "minVramGb": { "type": "number", "exclusiveMinimum": 0 },
+        "preferredGpuTypes": { "type": "array", "items": { "type": "string", "minLength": 1 } },
+        "cpuFallbackAllowed": { "type": "boolean" },
+        "preemptionSafe": { "type": "boolean" },
+        "estimatedDurationMinutes": {
+          "type": "object",
+          "additionalProperties": false,
+          "properties": {
+            "cpu": { "type": "number", "exclusiveMinimum": 0 },
+            "gpu": { "type": "number", "exclusiveMinimum": 0 }
+          }
+        },
+        "workloads": {
+          "description": "Exact reusable accounting batches with complete recipes, parameter/seed sets and total durations. Partition only for shared consumer subsets; one batch member does not require its own step or output object.",
+          "type": "array", "minItems": 1,
+          "items": { "type": "object", "required": ["id", "description", "estimatedDurationMinutes"], "additionalProperties": false,
+            "properties": { "id": { "type": "string", "minLength": 1 }, "description": { "type": "string", "minLength": 1 },
+              "estimatedDurationMinutes": { "type": "object", "minProperties": 1, "additionalProperties": false,
+                "properties": { "cpu": { "type": "number", "exclusiveMinimum": 0 }, "gpu": { "type": "number", "exclusiveMinimum": 0 } } }
+            }
+          }
+        },
+        "rationale": { "type": "string", "minLength": 1 }
+      }
+    }
+  }
+}