dsh-math-modeling-agent 0.3.1 → 0.5.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (43) hide show
  1. package/README.md +247 -172
  2. package/package.json +34 -34
  3. package/skills/math-modeling-agent/SKILL.md +70 -56
  4. package/skills/math-modeling-agent/references/claims-evidence.md +53 -41
  5. package/skills/math-modeling-agent/references/interaction-protocol.md +165 -163
  6. package/skills/math-modeling-agent/references/original-project-parity.md +19 -0
  7. package/skills/math-modeling-agent/references/report-contract.md +130 -122
  8. package/skills/math-modeling-agent/references/run-directory.md +65 -62
  9. package/skills/math-modeling-agent/references/subagent-dispatch.md +3 -2
  10. package/skills/math-modeling-agent/references/tool-policy.md +11 -0
  11. package/skills/math-modeling-agent/references/verification-recipes.md +43 -0
  12. package/skills/math-modeling-agent/references/workflow.md +10 -9
  13. package/skills/math-modeling-agent/schemas/attempt.schema.json +89 -70
  14. package/skills/math-modeling-agent/schemas/evidence.schema.json +110 -0
  15. package/skills/math-modeling-agent/schemas/failure.schema.json +30 -0
  16. package/skills/math-modeling-agent/schemas/ledger.schema.json +88 -68
  17. package/skills/math-modeling-agent/schemas/run.schema.json +126 -102
  18. package/skills/math-modeling-agent/schemas/verification.schema.json +60 -0
  19. package/skills/math-modeling-agent/scripts/computation/README.md +20 -0
  20. package/skills/math-modeling-agent/scripts/computation/backend-inventory.schema.json +78 -0
  21. package/skills/math-modeling-agent/scripts/computation/backend_inventory.ps1 +351 -0
  22. package/skills/math-modeling-agent/scripts/computation/backend_inventory.py +322 -0
  23. package/skills/math-modeling-agent/scripts/computation/computation_record.py +361 -0
  24. package/skills/math-modeling-agent/scripts/computation/probe_backends.ps1 +396 -0
  25. package/skills/math-modeling-agent/scripts/computation/probe_backends.py +230 -0
  26. package/skills/math-modeling-agent/scripts/correction-lineage.mjs +78 -0
  27. package/skills/math-modeling-agent/scripts/distribution-parity.mjs +123 -0
  28. package/skills/math-modeling-agent/scripts/evidence-store.mjs +349 -0
  29. package/skills/math-modeling-agent/scripts/failure-insights.mjs +79 -0
  30. package/skills/math-modeling-agent/scripts/input-snapshot.mjs +98 -0
  31. package/skills/math-modeling-agent/scripts/ledger-mutation.mjs +82 -0
  32. package/skills/math-modeling-agent/scripts/migration-v3.mjs +31 -0
  33. package/skills/math-modeling-agent/scripts/output-integrity.mjs +82 -0
  34. package/skills/math-modeling-agent/scripts/paper-evidence.mjs +45 -0
  35. package/skills/math-modeling-agent/scripts/report-contract.mjs +112 -0
  36. package/skills/math-modeling-agent/scripts/run-state.mjs +829 -707
  37. package/skills/math-modeling-agent/scripts/verification-recipes.mjs +52 -0
  38. package/skills/math-modeling-agent/scripts/verification-runner.mjs +8 -0
  39. package/skills/math-modeling-audit/SKILL.md +41 -41
  40. package/skills/math-modeling-audit/references/mcm-icm-final-judge.md +335 -331
  41. package/skills/math-modeling-audit/scripts/mcm-score.mjs +293 -218
  42. package/skills/math-modeling-audit/scripts/paper-final-review.mjs +40 -0
  43. package/skills/math-modeling-audit/scripts/project-initial-review.mjs +41 -0
package/skills/math-modeling-agent/schemas/attempt.schema.json CHANGED
@@ -1,70 +1,89 @@
1
- {
2
- "$schema": "https://json-schema.org/draft/2020-12/schema",
3
- "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/attempt.schema.json",
4
- "title": "Math Modeling Attempt",
5
- "type": "object",
6
- "additionalProperties": false,
7
- "required": [
8
- "schemaVersion",
9
- "attempt",
10
- "candidateId",
11
- "objective",
12
- "evidenceIds",
13
- "closedObligationIds",
14
- "issuesOpened",
15
- "issuesClosed",
16
- "progress",
17
- "nextAction",
18
- "status"
19
- ],
20
- "properties": {
21
- "schemaVersion": {
22
- "const": 1
23
- },
24
- "attempt": {
25
- "type": "integer",
26
- "minimum": 1
27
- },
28
- "candidateId": {
29
- "type": "string"
30
- },
31
- "objective": {
32
- "type": "string",
33
- "minLength": 1
34
- },
35
- "evidenceIds": {
36
- "type": "array",
37
- "items": {
38
- "type": "string"
39
- }
40
- },
41
- "closedObligationIds": {
42
- "type": "array",
43
- "items": {
44
- "type": "string"
45
- }
46
- },
47
- "issuesOpened": {
48
- "type": "array",
49
- "items": {
50
- "type": "string"
51
- }
52
- },
53
- "issuesClosed": {
54
- "type": "array",
55
- "items": {
56
- "type": "string"
57
- }
58
- },
59
- "progress": {
60
- "type": "boolean"
61
- },
62
- "nextAction": {
63
- "type": "string",
64
- "minLength": 1
65
- },
66
- "status": {
67
- "type": "string"
68
- }
69
- }
70
- }
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/attempt.schema.json",
4
+ "title": "Math Modeling Attempt",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": [
8
+ "schemaVersion",
9
+ "attempt",
10
+ "candidateId",
11
+ "objective",
12
+ "evidenceIds",
13
+ "closedObligationIds",
14
+ "issuesOpened",
15
+ "issuesClosed",
16
+ "progress",
17
+ "nextAction",
18
+ "status",
19
+ "scientificSections",
20
+ "coverage",
21
+ "symbols",
22
+ "verificationManifest"
23
+ ],
24
+ "properties": {
25
+ "schemaVersion": {
26
+ "const": 2
27
+ },
28
+ "attempt": {
29
+ "type": "integer",
30
+ "minimum": 1
31
+ },
32
+ "candidateId": {
33
+ "type": "string"
34
+ },
35
+ "objective": {
36
+ "type": "string",
37
+ "minLength": 1
38
+ },
39
+ "evidenceIds": {
40
+ "type": "array",
41
+ "items": {
42
+ "type": "string"
43
+ }
44
+ },
45
+ "closedObligationIds": {
46
+ "type": "array",
47
+ "items": {
48
+ "type": "string"
49
+ }
50
+ },
51
+ "issuesOpened": {
52
+ "type": "array",
53
+ "items": {
54
+ "type": "string"
55
+ }
56
+ },
57
+ "issuesClosed": {
58
+ "type": "array",
59
+ "items": {
60
+ "type": "string"
61
+ }
62
+ },
63
+ "progress": {
64
+ "type": "boolean"
65
+ },
66
+ "nextAction": {
67
+ "type": "string",
68
+ "minLength": 1
69
+ },
70
+ "status": {
71
+ "type": "string"
72
+ },
73
+ "scientificSections": {
74
+ "type": "array",
75
+ "minItems": 9,
76
+ "items": { "type": "string", "minLength": 1 }
77
+ },
78
+ "coverage": {
79
+ "type": "object"
80
+ },
81
+ "symbols": {
82
+ "type": "array",
83
+ "items": { "type": "object" }
84
+ },
85
+ "verificationManifest": {
86
+ "type": "object"
87
+ }
88
+ }
89
+ }
package/skills/math-modeling-agent/schemas/evidence.schema.json ADDED
@@ -0,0 +1,110 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/evidence.schema.json",
4
+ "title": "Math Modeling Evidence",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": [
8
+ "id",
9
+ "kind",
10
+ "coveredClaimIds",
11
+ "coveredObligationIds",
12
+ "artifactPath",
13
+ "inputHashes",
14
+ "outputHashes",
15
+ "command",
16
+ "workdir",
17
+ "environment",
18
+ "exitCode",
19
+ "methodFamily",
20
+ "codeHash",
21
+ "level",
22
+ "limitations"
23
+ ],
24
+ "properties": {
25
+ "id": {
26
+ "type": "string",
27
+ "minLength": 1
28
+ },
29
+ "kind": {
30
+ "type": "string",
31
+ "minLength": 1
32
+ },
33
+ "coveredClaimIds": {
34
+ "type": "array",
35
+ "minItems": 1,
36
+ "items": {
37
+ "type": "string",
38
+ "minLength": 1
39
+ }
40
+ },
41
+ "coveredObligationIds": {
42
+ "type": "array",
43
+ "minItems": 1,
44
+ "items": {
45
+ "type": "string",
46
+ "minLength": 1
47
+ }
48
+ },
49
+ "artifactPath": {
50
+ "type": "string",
51
+ "minLength": 1
52
+ },
53
+ "inputHashes": {
54
+ "type": "object",
55
+ "minProperties": 1,
56
+ "additionalProperties": {
57
+ "type": "string",
58
+ "minLength": 1
59
+ }
60
+ },
61
+ "outputHashes": {
62
+ "type": "object",
63
+ "minProperties": 1,
64
+ "additionalProperties": {
65
+ "type": "string",
66
+ "minLength": 1
67
+ }
68
+ },
69
+ "command": {
70
+ "type": "string",
71
+ "minLength": 1
72
+ },
73
+ "workdir": {
74
+ "type": "string",
75
+ "minLength": 1
76
+ },
77
+ "environment": {
78
+ "type": "object",
79
+ "minProperties": 1,
80
+ "additionalProperties": true
81
+ },
82
+ "exitCode": {
83
+ "type": "integer"
84
+ },
85
+ "methodFamily": {
86
+ "type": "string",
87
+ "minLength": 1
88
+ },
89
+ "codeHash": {
90
+ "type": "string",
91
+ "minLength": 1
92
+ },
93
+ "level": {
94
+ "enum": [
95
+ "DERIVED",
96
+ "EXECUTED",
97
+ "VERIFIED",
98
+ "INDEPENDENTLY_VERIFIED",
99
+ "EXTERNALLY_VALIDATED",
100
+ "LEGACY_UNVERIFIED"
101
+ ]
102
+ },
103
+ "limitations": {
104
+ "type": "array",
105
+ "items": {
106
+ "type": "string"
107
+ }
108
+ }
109
+ }
110
+ }
package/skills/math-modeling-agent/schemas/failure.schema.json ADDED
@@ -0,0 +1,30 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/failure.schema.json",
4
+ "title": "Math Modeling Failure",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": ["failureId", "stage", "subproblem", "attempt", "rootCause", "candidateId", "assumptionIds", "symptom", "minimalReproduction", "command", "inputHashes", "codeHash", "exitCode", "affectedClaims", "affectedObligations", "whatWasRuledOut", "scientificInsight", "recovery", "status", "evidenceIds"],
8
+ "properties": {
9
+ "failureId": { "type": "string", "minLength": 1 },
10
+ "stage": { "type": "string", "minLength": 1 },
11
+ "subproblem": { "type": "string", "minLength": 1 },
12
+ "attempt": { "type": "integer", "minimum": 1 },
13
+ "rootCause": { "enum": ["DATA", "TOOL", "MODEL", "ASSUMPTION", "ANALYSIS", "VALIDATION", "EVIDENCE", "RESEARCH_GAP"] },
14
+ "candidateId": { "type": "string", "minLength": 1 },
15
+ "assumptionIds": { "type": "array", "items": { "type": "string", "minLength": 1 } },
16
+ "symptom": { "type": "string", "minLength": 1 },
17
+ "minimalReproduction": { "type": "string", "minLength": 1 },
18
+ "command": { "type": "string", "minLength": 1 },
19
+ "inputHashes": { "type": "object", "minProperties": 1, "additionalProperties": { "type": "string", "minLength": 1 } },
20
+ "codeHash": { "type": "string", "minLength": 1 },
21
+ "exitCode": { "type": "integer" },
22
+ "affectedClaims": { "type": "array", "items": { "type": "string", "minLength": 1 } },
23
+ "affectedObligations": { "type": "array", "items": { "type": "string", "minLength": 1 } },
24
+ "whatWasRuledOut": { "type": "array", "items": { "type": "string", "minLength": 1 } },
25
+ "scientificInsight": { "type": "string", "minLength": 1 },
26
+ "recovery": { "type": "string", "minLength": 1 },
27
+ "status": { "enum": ["OPEN", "RECOVERED", "ABANDONED", "STALE"] },
28
+ "evidenceIds": { "type": "array", "items": { "type": "string", "minLength": 1 } }
29
+ }
30
+ }
package/skills/math-modeling-agent/schemas/ledger.schema.json CHANGED
@@ -1,68 +1,88 @@
1
- {
2
- "$schema": "https://json-schema.org/draft/2020-12/schema",
3
- "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/ledger.schema.json",
4
- "title": "Math Modeling Ledger",
5
- "type": "object",
6
- "additionalProperties": false,
7
- "required": [
8
- "schemaVersion",
9
- "taskId",
10
- "scope",
11
- "assumptions",
12
- "claims",
13
- "obligations",
14
- "subproblems",
15
- "candidates",
16
- "issues"
17
- ],
18
- "properties": {
19
- "schemaVersion": {
20
- "const": 2
21
- },
22
- "taskId": {
23
- "type": "string"
24
- },
25
- "scope": {
26
- "type": "object",
27
- "properties": {
28
- "independentAuditPassed": { "type": "boolean" },
29
- "interactions": {
30
- "type": "array",
31
- "description": "L1 decision-point records: {id, decisionPoint: D0|D1|D-R|D2|D3|D4, question, options, userAnswer, timestamp, effectOnModel} or {auto: true, decisionPoint, reason}; D-R is the pre-modeling literature-research exchange",
32
- "items": { "type": "object" }
33
- },
34
- "decisionStack": {
35
- "type": "array",
36
- "description": "Decision-point stack for backtracking: {id (e.g. D3-SP1), point, choice, at, reason}. REVISE/RESEARCH reasons must reference an entry id.",
37
- "items": { "type": "object" }
38
- },
39
- "robustnessExempt": {
40
- "type": "boolean",
41
- "description": "True only for tasks with no perturbable parameters (pure proofs); terminal states otherwise require a PASSED robustness-kind obligation."
42
- }
43
- }
44
- },
45
- "assumptions": {
46
- "type": "array",
47
- "description": "Per-subproblem incremental assumptions: {id, statement, scope: [SPx]|global, source, status: active|revised|superseded, risk, sensitivityPlan, revisionHistory: [{at, reason, changedTo}]}",
48
- "items": { "type": "object" }
49
- },
50
- "claims": {
51
- "type": "array"
52
- },
53
- "obligations": {
54
- "type": "array"
55
- },
56
- "subproblems": {
57
- "type": "array",
58
- "description": "Subproblem DAG nodes: {id, task, inputs, dependencies, acceptance, status, analysisLog: [{at, entry, trigger}]} — analysisLog is the rolling analysis ledger, updated after every deep-exchange beat.",
59
- "items": { "type": "object" }
60
- },
61
- "candidates": {
62
- "type": "array"
63
- },
64
- "issues": {
65
- "type": "array"
66
- }
67
- }
68
- }
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://github.com/yohanchen1/MathModelingAgent/schemas/ledger.schema.json",
4
+ "title": "Math Modeling Ledger",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": [
8
+ "schemaVersion",
9
+ "taskId",
10
+ "scope",
11
+ "assumptions",
12
+ "claims",
13
+ "obligations",
14
+ "subproblems",
15
+ "candidates",
16
+ "issues",
17
+ "requirements",
18
+ "evidence",
19
+ "verifications",
20
+ "failures"
21
+ ],
22
+ "properties": {
23
+ "schemaVersion": {
24
+ "const": 3
25
+ },
26
+ "taskId": {
27
+ "type": "string"
28
+ },
29
+ "scope": {
30
+ "type": "object",
31
+ "properties": {
32
+ "independentAuditPassed": { "type": "boolean" },
33
+ "interactions": {
34
+ "type": "array",
35
+ "description": "L1 decision-point records: {id, decisionPoint: D0|D1|D-R|D2|D3|D4, question, options, userAnswer, timestamp, effectOnModel} or {auto: true, decisionPoint, reason}; D-R is the pre-modeling literature-research exchange",
36
+ "items": { "type": "object" }
37
+ },
38
+ "decisionStack": {
39
+ "type": "array",
40
+ "description": "Decision-point stack for backtracking: {id (e.g. D3-SP1), point, choice, at, reason}. REVISE/RESEARCH reasons must reference an entry id.",
41
+ "items": { "type": "object" }
42
+ },
43
+ "robustnessExempt": {
44
+ "type": "boolean",
45
+ "description": "True only for tasks with no perturbable parameters (pure proofs); terminal states otherwise require a PASSED robustness-kind obligation."
46
+ }
47
+ }
48
+ },
49
+ "assumptions": {
50
+ "type": "array",
51
+ "description": "Per-subproblem incremental assumptions: {id, statement, scope: [SPx]|global, source, status: active|revised|superseded, risk, sensitivityPlan, revisionHistory: [{at, reason, changedTo}]}",
52
+ "items": { "type": "object" }
53
+ },
54
+ "claims": {
55
+ "type": "array"
56
+ },
57
+ "obligations": {
58
+ "type": "array"
59
+ },
60
+ "subproblems": {
61
+ "type": "array",
62
+ "description": "Subproblem DAG nodes: {id, task, inputs, dependencies, acceptance, status, analysisLog: [{at, entry, trigger}]} — analysisLog is the rolling analysis ledger, updated after every deep-exchange beat.",
63
+ "items": { "type": "object" }
64
+ },
65
+ "candidates": {
66
+ "type": "array"
67
+ },
68
+ "issues": {
69
+ "type": "array"
70
+ },
71
+ "requirements": {
72
+ "type": "array",
73
+ "items": { "type": "object" }
74
+ },
75
+ "evidence": {
76
+ "type": "array",
77
+ "items": { "type": "object" }
78
+ },
79
+ "verifications": {
80
+ "type": "array",
81
+ "items": { "type": "object" }
82
+ },
83
+ "failures": {
84
+ "type": "array",
85
+ "items": { "type": "object" }
86
+ }
87
+ }
88
+ }