open-problem-lab/schemas/task.schema.json at main · Open-Problem-Lab/open-problem-lab · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://open-problem-lab.local/schemas/task.schema.json",
  "title": "Task",
  "description": "An atomic unit of work within a problem pack. A task must have a binary done condition — it either holds or it does not. Tasks that cannot fail have not been scoped.",
  "type": "object",
  "additionalProperties": false,
  "required": [
    "id",
    "problem_id",
    "title",
    "status",
    "owner_role",
    "outcome",
    "inputs",
    "expected_artifact",
    "validation_method",
    "reviewer_needed",
    "safety_risk",
    "done_condition"
  ],
  "properties": {
    "id": {
      "type": "string",
      "pattern": "^[a-z0-9]+(?:-[a-z0-9]+)*$",
      "description": "Unique task identifier in kebab-case. Referenced in agent submissions and issue titles. Example: source-inventory"
    },
    "problem_id": {
      "type": "string",
      "pattern": "^[a-z0-9]+(?:-[a-z0-9]+)*/[a-z0-9]+(?:-[a-z0-9]+)*$",
      "description": "The problem pack this task belongs to. Must match the id field in problem.json. Example: climate-health/malaria-early-warning-africa"
    },
    "title": {
      "type": "string",
      "minLength": 8,
      "description": "One sentence, action-oriented. The title should make clear what is being done and to what. Example: 'Inventory malaria and climate data sources for Sub-Saharan Africa'"
    },
    "status": {
      "type": "string",
      "description": "Current workflow state. Only tasks with status 'scoped' are ready for contributors to pick up.",
      "enum": [
        "needs-triage",
        "scoped",
        "agent-working",
        "needs-review",
        "needs-replication",
        "accepted",
        "field-tested",
        "rejected"
      ]
    },
    "owner_role": {
      "type": "string",
      "description": "The agent role guide best suited for this task. Contributors should read that guide before starting. See agents/ directory.",
      "enum": [
        "literature-scout",
        "data-cleaner",
        "red-team-reviewer",
        "field-reality-reviewer",
        "implementation-planner"
      ]
    },
    "outcome": {
      "type": "string",
      "minLength": 20,
      "description": "What exists when this task is complete. State the artifact or decision, not the effort. Example: 'A dated source inventory separating usable, limited, and rejected sources with explicit reasons for each classification.'"
    },
    "inputs": {
      "type": "array",
      "description": "Specific data sources, prior artifacts, or documents needed to complete this task. Name them precisely enough that a contributor unfamiliar with the pack can find them.",
      "items": {
        "type": "string",
        "minLength": 3
      },
      "minItems": 1
    },
    "expected_artifact": {
      "type": "string",
      "minLength": 8,
      "description": "The specific file or section that changes when this task is done. Example: 'datasets.md update plus evidence.json records for each accepted source'"
    },
    "validation_method": {
      "type": "string",
      "minLength": 10,
      "description": "How a reviewer or replicator checks that the done condition holds. Must produce a verifiable outcome — not 'reviewer confirms the analysis is correct.' Example: 'Reviewer checks each source for URL, date, access status, and stated limitations.'"
    },
    "reviewer_needed": {
      "type": "string",
      "description": "Role of reviewer required. Use one of: domain-reviewer, replicator, red-team-reviewer, field-reality-reviewer. Example: 'domain-reviewer'",
      "enum": [
        "domain-reviewer",
        "replicator",
        "red-team-reviewer",
        "field-reality-reviewer"
      ]
    },
    "safety_risk": {
      "type": "string",
      "description": "Risk level for this task's output. low = framing, literature summaries. medium = data cleaning, ranking, models. high = operational recommendations, forecasts, resource allocation.",
      "enum": ["low", "medium", "high"]
    },
    "done_condition": {
      "type": "string",
      "minLength": 20,
      "description": "A single falsifiable statement that holds when the task is complete. If the done condition is always reachable, the task has not been scoped. Example: 'At least five candidate data sources are classified as usable, limited, or rejected with explicit reasons.'"
    }
  }
}