dsh-math-modeling-agent 0.2.8 → 0.4.1

package/README.md

@@ -71,6 +71,10 @@ NOT_CHECKED → DERIVED → EXECUTED → VERIFIED → INDEPENDENTLY_VERIFIED →
 
 工具只是产生证据的方式，可以替换，工作流不变。
 
+本插件保留原 GitHub 项目的核心行为：问题/附件探查、子问题顺序求解、Modeler → Analyzer → Correction 迭代、逐轮日志和可恢复运行；对应关系见 `skills/math-modeling-agent/references/original-project-parity.md`。
+
+数学计算后端扩展位于 `skills/math-modeling-agent/scripts/computation/`：它只负责探测可调用后端、生成能力快照和校验计算记录，不捆绑 Mathematica/SageMath 等外部引擎。
+
 - **Python** 可选（推荐）：需要计算时在运行目录内创建隔离环境（数值计算、数据分析、优化、仿真、绘图、独立重算）
 - **Lean** 可选：形式化验证；不自动安装；**形式命题被证明 ≠ 现实主张被证明**
 - **Wolfram** 可选：符号计算、解析推导、恒等式验证
@@ -108,7 +112,7 @@ D3 方向/D4 裁决），`run-state.mjs gate` 在每次状态转移前强制校
 安装：
 
 ```bash
-dsh plugin --profile web add github:yohanchen1/MathModelingAgent#v0.2.8
+dsh plugin --profile web add github:yohanchen1/MathModelingAgent#v0.4.1
 dsh --profile web --dump-config   # 检查组合层（应看到 dsh-math-modeling-agent-skills 行）
 dsh web                          # 重启以加载插件
 ```
@@ -118,6 +122,17 @@ dsh web                          # 重启以加载插件
 不要用 `npm install` 代替，那会装到错误位置。如本机镜像源同步滞后，可显式走官方源：
 `dsh plugin --profile web add dsh-math-modeling-agent --registry=https://registry.npmjs.org/`）
 
+维护者在发布后应分别核对 GitHub 解包目录、npm tarball 解包目录和 profile
+安装目录：
+
+```bash
+node skills/math-modeling-agent/scripts/distribution-parity.mjs <source-root> <candidate-root>
+```
+
+该命令比较 package.json `files` allowlist 内所有文件的 SHA-256；它不把 AI
+解题中的偶发失败当作分发失败。`dsh --profile web --dump-config` 用于检查组合，
+但新版 DSH 可能重写 profile 的空 `cordis.yml`，不是严格只读操作。
+
 开始建模——直接描述任务即可。每个环节都是人与 LLM 的深度信息交换：LLM 给出带依据链的
 完整分析（题面原句/数据证据/文献/显式判断标记），你纠正、补充背景或提供自己的参考文献，
 LLM 更新并展示差异。建模前必须先做文献调研（AI 检索原理与方法文献，你可增删），每个候选

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dsh-math-modeling-agent",
-  "version": "0.2.8",
+  "version": "0.4.1",
   "description": "Evidence-driven mathematical modeling and verification skills for DeepSeek Harness",
   "type": "module",
   "files": [

package/skills/math-modeling-agent/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: math-modeling-agent
-description: This skill should be used when the user asks to "建立数学模型", "解决数学建模题", "继续上次建模", "寻找新的建模方向", "solve this modeling problem", or needs an evidence-backed mathematical model with computation, research, verification, per-attempt reports, and resumable state.
-whenToUse: Use for open-ended mathematical modeling, prediction, optimization, estimation, simulation, mechanism, decision, or multi-part contest problems; do not use for auditing an already completed artifact without changing it.
+description: 当用户要求分析数学建模题、读取题目附件、建立预测/优化/估计/仿真模型、比较建模方向、执行数值验证、进行鲁棒性分析或继续已有建模项目时使用。
+whenToUse: 用于开放式数学建模、预测、优化、估计、仿真、机制分析、决策和多小问竞赛题；已有成果的独立审计改用 math-modeling-audit，不在此处静默改写被审计成果。
 user-invocable: true
 ---
 
@@ -17,23 +17,23 @@ Accept problem text, attachments, or an existing run directory.
 
 Interaction model (references/interaction-protocol.md): execution details stay autonomous (file names, tool adapters, state fields), but every model-shaping decision is a DEEP three-beat exchange — LLM presents a complete analysis with its evidence chain, the user responds with substance (corrections, background, their own references), the LLM updates the artifact and shows the delta. Decision points: D0 six-section restatement / D1 routing once per task, D-R literature research before any candidate is built, D2 assumptions and D3 direction per subproblem, D2' on mid-subproblem discoveries, D4 verdict per round. Show the runlog digest in the chat after every round (references/report-contract.md); announce one status line before every transition; the user may interrupt at any time.
 
-Return a concise current status, supported claims, unresolved obligations, attempt delta, artifact path, and next action or final report.
+Return a concise current status, supported claims, unresolved obligations, attempt delta, artifact path, and next action or final report. The legacy Modeler → Analyzer → Correction behavior mapping is recorded in `references/original-project-parity.md`.
 
 ## Default workflow
 
 1. Route the task to Fast, Standard, or High-Assurance using `references/problem-types.md`; ask D1 (task type + assurance mode).
-2. Read available inputs before asking questions; apply the per-subproblem math grill in `references/math-grill.md`; show the D0 restatement and intake digest in the chat.
-3. Create or validate run state with `scripts/run-state.mjs` and follow `references/state-recovery.md`. Gate EVERY transition: `run-state.mjs gate <dir> --to <status> --reason "..."`.
-4. Profile inputs and build the subproblem DAG using `references/data-subproblems.md`.
-5. Register assumptions (per subproblem), claims, and verification obligations from `references/claims-evidence.md`; ask D2 for the subproblem being started.
-6. **Literature research (mandatory, before candidates)**: transition CLAIMS_REGISTERED → RESEARCH, search principles and methods by direction (never contest solutions), write `research/sources.jsonl`, attach a literature column to the candidate matrix, then ask D-R (user endorses/rejects/adds references) and move to CANDIDATES_READY.
-7. Build a baseline and materially different candidates using `references/modeling-methodology.md`; ask D3 (candidate matrix).
-8. Probe tools with `scripts/capability-probe.mjs`; create a run-local Python environment with `scripts/python-environment.mjs` only when computation is required. Announce the tool choice and fallback for the phase (references/tool-policy.md).
-9. Execute, verify, critique, and revise according to `references/workflow.md` and `references/tool-policy.md`; ask D4 after each VERIFY.
+2. Literature research FIRST: transition TRIAGE → RESEARCH, search principles and methods by direction (never contest solutions), write `research/sources.jsonl` + methods matrix, ask D-R (user endorses/rejects/adds).
+3. Show the D0 deep restatement — six sections plus the question-dependency relation — and ask D0; gate SCOPE_FROZEN (requires D0+D1+D-R).
+4. Panorama analysis + data profile (INPUT_PROFILED) with user interaction; global assumptions G-class confirmed via D2-G before CLAIMS_REGISTERED.
+5. Register claims and obligations; build the candidate matrix with literature column (CANDIDATES_READY) and ask D3.
+6. For EACH subproblem: subproblem analysis (reference the panorama, do not re-derive), S-class assumptions via D2-SPi, then ATTEMPT → EXECUTE → VERIFY → subproblem robustness (plan→interact→run) → D4 verdict → REVISE or move to the next subproblem.
+7. Live documents: any discovery that shows an earlier phase was incomplete patches that phase incrementally (revision + stale), never re-running unaffected completed work.
+8. Backtracking depth follows the failure root cause (attempt/verify/direction/assumption/analysis/restatement/research), recorded in failed/issues.md.
+9. Probe tools with `scripts/capability-probe.mjs`; create a run-local Python environment with `scripts/python-environment.mjs` when computation is required; announce the tool choice and fallback.
 10. Dispatch subagents for parallel exploration, blind verification, and audits per `references/subagent-dispatch.md`.
-11. Escalate evidence gaps or stalled directions through `references/research-breakthrough.md`; write wall memos and the failure ledger per `references/run-directory.md`.
-12. Write only incremental attempt and terminal reports defined by `references/report-contract.md`; output the per-round runlog digest in the chat.
-13. Before any terminal transition, run the cleanup gate (references/run-directory.md) and the robustness gate (a PASSED robustness-kind obligation, or scope.robustnessExempt with reason).
+11. Escalate evidence gaps through `references/research-breakthrough.md`; write wall memos and the failure ledger per `references/run-directory.md`.
+12. Write incremental attempt and terminal reports per `references/report-contract.md`; output the per-round runlog digest in the chat.
+13. Before any terminal transition: cleanup gate, ALL required robustness obligations PASS (or robustnessExempt), every subproblem DONE with dependencies satisfied, D4 for the current attempt, report.md present.
 
 ## Invariants
 
@@ -47,6 +47,19 @@ Return a concise current status, supported claims, unresolved obligations, attem
 - Continue only when a round closes an obligation, adds reproducible evidence, removes a blocker, tightens uncertainty, or refutes a candidate.
 - The run directory stays clean at every moment, and every failed path is preserved in its own folder (references/run-directory.md).
 
+## Computation tool extension
+
+When a computation is required, first call the bundled backend probe:
+`python scripts/computation/probe_backends.py`
+(or the `.ps1` entrypoint when Python is unavailable). Choose only a backend
+reported callable; record the primary route and fallback. Use
+`scripts/computation/backend_inventory.py` with a run-local `--state-file`
+when a persistent capability snapshot is needed. For reproducible/file or
+high-assurance work, initialize and validate
+`scripts/computation/computation_record.py`. These scripts discover and record
+external backends; they are not mathematical engines and their successful
+execution is not proof.
+
 ## Tool degradation
 
 Python is recommended but not required for installation. Lean and Wolfram are optional. If a required tool is unavailable, weaken the claim and report the missing obligation; never pretend execution or formal verification occurred.

package/skills/math-modeling-agent/references/claims-evidence.md

@@ -6,7 +6,9 @@ Store claim ID, exact wording, type, scope/quantifiers, assumptions, risk, verif
 
 ## Assumption record (v2)
 
-Store id, statement, scope (subproblem ids or global), source category, risk, sensitivity plan, validation status, affected claims, status (active/revised/superseded), and revisionHistory [{at, reason, changedTo}]. Allowed source categories are problem, data, theory, verified literature, domain instruction, and modeling simplification.
+Store id, statement, scope ('global' for whole-problem assumptions, or a subproblem id list), source category, risk, sensitivity plan, validation status, affected claims, status (active/revised/superseded), and revisionHistory [{at, reason, changedTo}]. Allowed source categories are problem, data, theory, verified literature, domain instruction, and modeling simplification.
+
+Two classes: G-class (global, confirmed once in the panorama layer via D2-G) and S-class (per subproblem, confirmed via D2-SPi). Claims carry an `assumptions` field listing the assumption ids they depend on; when an assumption is revised, dependent claims are marked stale and must carry post-revision evidence before a terminal state (validate warns otherwise).
 
 Every assumption additionally carries a `basis` field — the evidence chain behind it: `problem-quote` (exact original wording), `data` (evidence id + statistic), `literature` (source id + excerpt location), or `llm-judgment` (explicitly flagged as the LLM's own inference, challengeable in D2).
 

package/skills/math-modeling-agent/references/interaction-protocol.md

@@ -31,7 +31,8 @@ until the user says "可以了". Mathematical modeling is a deep activity;
 | D0 问题重述 | intake complete | the six-section deep restatement below + open clarification questions | accept + no additions |
 | D1 路由 | TRIAGE | task type (prediction/optimization/explanation/proof/…) + assurance mode (Fast/Standard/High-Assurance) | auto-routing result |
 | D-R 文献研究 | CLAIMS_REGISTERED (before CANDIDATES_READY) | the literature list by direction (AI-found); user endorses / rejects / adds their own | accept list |
-| D2 假设确认 | **each subproblem starts** | the assumptions for THIS subproblem, each with statement + basis chain (problem quote / data evidence / literature / LLM-judgment marker) + risk + challengeable point | accept all |
+| D2-G 全局假设确认 | CLAIMS_REGISTERED (全景层) | the GLOBAL assumptions that hold for the whole problem (geometry, data, physical mechanism, numerical basis) — each with basis chain + risk | accept all |
+| D2 子问题假设确认 | **each subproblem starts** | the assumptions for THIS subproblem only (S-class), each with statement + basis chain + risk | accept all |
 | D2' 新假设插问 | mid-subproblem discovery that changes model structure | the new assumption + its effect | record-and-continue |
 | D3 方向选择 | CANDIDATES_READY | candidate matrix with a literature column per direction; user picks or authorizes auto | baseline |
 | D4 结果裁决 | VERIFY done | verdict + evidence digest + key intermediate results (residual/constraint-violation distributions, figure paths) + failure details | verdict-implied |
@@ -49,6 +50,9 @@ until the user says "可以了". Mathematical modeling is a deep activity;
 5. **各子问题数学本质判定**：every sub-question of the problem — however
    many there are (one or ten) — is classified (fitting/optimization/
    geometric optics/simulation/…), with why.
+5.5 **问间依赖关系**：which sub-questions depend on which (models, data,
+   verified claims reused downstream) — the整体感 guarantee: later questions
+   REFERENCE earlier verified results instead of re-deriving them.
 6. **待澄清问题**：open questions inviting the user to add background,
    correct a reading, or supply their own references — never a yes/no list.
 
@@ -141,6 +145,14 @@ interruption opportunity after the first step.
   and every downstream claim/obligation that depended on the old assumption is
   marked `stale: true` with the revising assumption id (see claims-evidence.md).
 
+## Live documents (every phase can be refined later)
+
+Every phase artifact is LIVE: problem-brief.md (restatement), analysisLog,
+assumptions, sources.jsonl, methods.md, attempt reports. While solving,
+anything discovered that shows an earlier phase was incomplete is patched
+INCREMENTALLY — revision record + downstream stale marking — without
+re-running completed unaffected work. This is the "解题中不断完善" rule.
+
 ## Rolling documents
 
 - Problem restatement: written once at intake (problem-brief.md); only

package/skills/math-modeling-agent/references/math-grill.md

@@ -17,6 +17,14 @@ Ask one question at a time. Read the problem and attachments first. Ask only que
 10. 假设及依据；11. 任务类型（预测/优化/解释/证明/机制/决策）；12. 误差/精度要求；
 13. 验证标准；14. 最终交付物；15. 风险与失败模式；16. 不适用字段（标注原因）。
 
+## Layered questioning (v0.3)
+
+- **全景层**：global assumptions (G-class) are grilled once — geometry, data
+  quality, physical mechanism, numerical basis; recorded via D2-G before
+  CLAIMS_REGISTERED.
+- **逐问层**：per-subproblem assumptions (S-class) are grilled when the
+  subproblem starts (D2-SPi) — only what is specific to this sub-question.
+
 ## Per-subproblem grilling (v2)
 
 - Grill **one subproblem at a time**. Never demand assumptions for Q2 while Q1 is unsolved.

package/skills/math-modeling-agent/references/original-project-parity.md

@@ -0,0 +1,19 @@
+# Original GitHub workflow parity
+
+The DSH bundle is a faithful workflow migration of the original
+MathModelingAgent project, not a replacement of its useful behavior.
+
+| Original GitHub behavior | DSH realization |
+|---|---|
+| PDF/problem extraction in `code/agent.py` | intake D0 and the subproblem DAG |
+| Excel attachment profiling in `code/data_analyzer.py` | INPUT_PROFILED and the run input manifest |
+| sequential subproblem solving | dependency-ordered SPi loop |
+| Modeler → Analyzer → Correction loop in `code/prompts.py` | explorer/modeler → verifier/analyzer → corrector/revise |
+| per-round thinking and run logs | attempts/<n>/report.md, events.jsonl, and chat runlog digest |
+| acceptance/rejection and stagnation control | D4 verdict, progress test, REVISE/FORK, and resumable non-SOLVED states |
+| hash-based cache and resume | run-local artifacts, evidence links, and state recovery |
+
+The migration deliberately replaces private chain-of-thought storage with
+human-readable reasoning, evidence, and reproducible artifacts. The roles and
+loop semantics remain; only the execution substrate changes from the legacy
+Python CLI to DSH skills and state files.

package/skills/math-modeling-agent/references/report-contract.md

@@ -24,6 +24,12 @@ Do not copy full prior solutions or private reasoning.
 
 State the frozen question, final scientific status, answer/recommendation, verified claims and evidence links, conditional/unresolved claims, assumptions and sensitivity, candidate comparison, data and parameter sources, validation, failure cases, limitations, reproducibility commands, and resume conditions when non-SOLVED.
 
+整体感 requirement (v0.3): the terminal report is ONE narrative, not a
+concatenation of sub-question reports — it must show the question-dependency
+chain (which model fed which), the shared global assumptions, and the
+cross-sub-question consistency checks. A reader must be able to see the whole
+problem as one connected argument.
+
 ## Chat display — the runlog digest (MUST output, v2.7)
 
 The following is shown DIRECTLY IN THE CHAT, not only on disk.

package/skills/math-modeling-agent/references/subagent-dispatch.md

@@ -7,8 +7,9 @@ produce reports and evidence. Never let a subagent mutate run-state directly.
 
 | Class | Role (lineage) | When | Count |
 |---|---|---|---|
-| explorer | parallel candidate exploration (legacy Modeler) | CANDIDATES_READY, >= 2 independent directions | 1 per direction |
-| verifier | blind re-verification (legacy Analyzer/Correction) | before terminal: key numeric/feasibility claims | Standard: 1 per task; High-Assurance: 1 per claim |
+| explorer | parallel candidate exploration (original Modeler lineage) | CANDIDATES_READY, >= 2 independent directions | 1 per direction |
+| verifier | blind re-verification (original Analyzer lineage) | before terminal: key numeric/feasibility claims | Standard: 1 per task; High-Assurance: 1 per claim |
+| corrector | apply accepted critique and produce the next attempt (original Correction lineage) | after a failed/conditional verdict | 1 per selected direction |
 | auditor | artifact-only audit (math-modeling-audit skill) | before SOLVED | High-Assurance: mandatory; else optional |
 
 ## Brief template (self-contained; subagents see no parent context)

package/skills/math-modeling-agent/references/tool-policy.md

@@ -22,6 +22,17 @@ e.g. "本阶段用 Python+OSQP 做约束优化；Wolfram 对本题非必需，
 The problem never silently depends on a single tool: every phase names its
 backend, its fallback, and the degradation path when the backend fails.
 
+## Computation backend extension
+
+Before selecting an external computation backend, run
+`scripts/computation/probe_backends.py` (or its PowerShell compatibility entry)
+and use only the live capability result. Persist a backend snapshot with
+`scripts/computation/backend_inventory.py` and an explicit run-local
+`--state-file` when the phase needs a stable inventory. Use
+`scripts/computation/computation_record.py` for reproducible/file or
+high-assurance computation. A probe, inventory, or record validation does not
+prove a mathematical claim.
+
 ## Python
 
 Use `scripts/python-environment.mjs`. Prefer uv, then an existing Python runtime plus venv. Keep the environment under the run directory. Install only required normalized PyPI package names; VCS URLs, local wheels, or arbitrary indexes require explicit user approval. Record lock, seed, command, workdir, timeout, stdout/stderr, exit code, and generated files.

package/skills/math-modeling-agent/references/workflow.md

@@ -8,21 +8,28 @@
 
 ## States
 
-TRIAGE → SCOPE_FROZEN → INPUT_PROFILED → CLAIMS_REGISTERED → CANDIDATES_READY → ATTEMPT → EXECUTE → VERIFY.
+TRIAGE → RESEARCH → SCOPE_FROZEN → INPUT_PROFILED → CLAIMS_REGISTERED → CANDIDATES_READY → ATTEMPT → EXECUTE → VERIFY.
 
 VERIFY may move to REVISE, RESEARCH, FORK, an independent audit, or one terminal status:
 SOLVED, PARTIAL, CONDITIONAL, INCONCLUSIVE, REFUTED, INFEASIBLE, UNIDENTIFIABLE, BLOCKED, CANCELLED.
 
-## Pre-modeling literature research (v2.6, mandatory)
+## Three-layer flow (v0.3)
 
-CLAIMS_REGISTERED → RESEARCH → CANDIDATES_READY. Before any candidate is
-built, run the D-R exchange (interaction-protocol.md): search principles and
-methods by direction, write research/sources.jsonl, attach a literature column
-to the candidate matrix. Every direction must be traceable to at least one
-source — engineering background, methodology paper, textbook, or official
-documentation; a literature-free direction must be explicitly marked
+- **全景层**（一次建立，贯穿全题）：TRIAGE(读题+D1) → RESEARCH(文献+D-R，前置) → SCOPE_FROZEN(D0 七段重述，含问间依赖) → INPUT_PROFILED(全景分析+数据画像) → CLAIMS_REGISTERED(全局假设 G，D2-G) → CANDIDATES_READY(方法矩阵+D3)。
+- **逐问层**（SPi 循环）：子问题分析 → 子问题假设 Si（D2-SPi）→ ATTEMPT 建模求解 → VERIFY 该问验证 → 该问鲁棒性（扰动计划→交互→执行）→ D4 裁决 → REVISE（本问修订）或进入 SPi+1（reason: "SPi 完成，进入 SPi+1"）。
+- **收束层**：跨问联动检查（终态 gate 校验全部子问题 DONE + 依赖满足）→ 全题鲁棒性汇总 → 终态硬门 → SOLVED + final-report（整体叙事，非拼凑）。
+
+## Pre-modeling literature research (mandatory)
+
+TRIAGE → RESEARCH → D-R → SCOPE_FROZEN. After the initial task triage and
+before D0 restatement is frozen, run the D-R exchange
+(interaction-protocol.md): search principles, engineering background, and
+methods by direction; write research/sources.jsonl; attach a literature
+column to the candidate matrix. Every direction must be traceable to at least
+one source. A literature-free direction must be explicitly marked
 "原创方向" with its derivation basis. Never search for solutions to the
-specific contest problem.
+specific contest problem. Mid-solve evidence gaps use a recorded
+RESEARCH backtrack, not an implicit bypass of the initial research phase.
 
 ## Interaction gates (v2)
 
@@ -37,6 +44,22 @@ scope.robustnessExempt is set with a reason.
 
 - REVISE/RESEARCH reasons MUST reference a decisionStack entry id; the round
   returns to that decision point and tries the next option.
+- Backtracking depth is chosen by the failure root cause (recorded in
+  failed/issues.md):
+
+| Root cause | Depth | Action |
+|---|---|---|
+| IMPLEMENTATION / TOOL | attempt | fix and rerun, no backtrack |
+| VALIDATION | verify | add obligations, re-verify |
+| MODEL (wrong direction) | direction | switch candidate (wall memo) |
+| assumption falsified | assumption | revise assumption (revisionHistory) → dependent claims stale → re-verify |
+| analysis gap | analysis | update analysisLog → re-derive downstream |
+| restatement misunderstanding | restatement | revise problem-brief (re-freeze) |
+| RESEARCH_GAP | research | add sources, switch method |
+
+  Backtracking to assumption/analysis/restatement depth is a CONTENT revision
+  plus downstream re-verification — the state machine does not rewind; the
+  loop is expressed as REVISE cycles and event records.
 - FORK records its fork point; each failed branch writes a wall memo
   (run-directory.md) before the next branch starts. A wrong direction always
   returns to the fork point, never mid-branch.

package/skills/math-modeling-agent/scripts/computation/README.md

@@ -0,0 +1,20 @@
+# DSH computation tool extension
+
+These entrypoints are adapted from the user-provided
+`math-science-computation` toolkit. They are computation-support tools, not
+mathematical engines: they discover callable external backends, persist a
+local capability snapshot, and validate reproducible computation records.
+
+## Entrypoints
+
+- `probe_backends.py` / `probe_backends.ps1`: read-only discovery of
+  Wolfram/Mathematica, primecount, SageMath, Python, and Python library
+  availability. External probes are bounded and return structured failures.
+- `backend_inventory.py` / `backend_inventory.ps1`: read or refresh a
+  capability snapshot. Pass a run-local `--state-file`/`-StateFile`; do not
+  use an implicit shared machine-wide path.
+- `computation_record.py`: initialize or validate a reproducibility record
+  for every probeable backend, including primecount.
+
+The external backend remains responsible for the actual mathematical
+calculation. A successful probe or record validation is not a proof.

package/skills/math-modeling-agent/scripts/computation/backend-inventory.schema.json

@@ -0,0 +1,78 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://local.codex.invalid/math-science-computation/backend-inventory.schema.json",
+  "title": "Math and science computation local backend inventory",
+  "type": "object",
+  "required": [
+    "inventory_schema_version",
+    "created_at_utc",
+    "updated_at_utc",
+    "local",
+    "mcp",
+    "invalidations"
+  ],
+  "properties": {
+    "inventory_schema_version": { "const": "1.0" },
+    "created_at_utc": { "type": "string", "format": "date-time" },
+    "updated_at_utc": { "type": "string", "format": "date-time" },
+    "local": {
+      "type": "object",
+      "required": ["schema_version", "probed_at_utc", "host", "mathematica", "primecount", "sagemath", "python"],
+      "properties": {
+        "schema_version": { "const": "1.0" },
+        "probed_at_utc": { "type": "string", "format": "date-time" },
+        "host": {
+          "type": "object",
+          "required": ["system", "architecture"],
+          "properties": {
+            "system": { "enum": ["Windows", "Darwin", "Linux"] },
+            "architecture": { "type": "string", "minLength": 1 }
+          },
+          "additionalProperties": true
+        },
+        "mathematica": { "type": "object" },
+        "primecount": { "type": "object" },
+        "sagemath": { "type": "object" },
+        "python": { "type": "object" }
+      },
+      "additionalProperties": true
+    },
+    "mcp": {
+      "type": "object",
+      "required": ["authority", "persisted_status", "required_action"],
+      "properties": {
+        "authority": { "const": "current_session_tool_discovery_and_call" },
+        "persisted_status": { "const": "historical_only" },
+        "required_action": { "type": "string" },
+        "mathematica": {
+          "type": "object",
+          "required": ["server_name", "protocol_version", "server_version", "wolfram_language_version", "observed_at_utc", "evidence"],
+          "properties": {
+            "server_name": { "type": "string", "minLength": 1 },
+            "protocol_version": { "type": "string", "pattern": "^[0-9]{4}-[0-9]{2}-[0-9]{2}$" },
+            "server_version": { "type": "string", "minLength": 1 },
+            "wolfram_language_version": { "type": "string", "minLength": 1 },
+            "observed_at_utc": { "type": "string", "format": "date-time" },
+            "evidence": { "const": "initialize_handshake_and_evaluator" }
+          },
+          "additionalProperties": false
+        }
+      },
+      "additionalProperties": false
+    },
+    "invalidations": {
+      "type": "array",
+      "items": {
+        "type": "object",
+        "required": ["backend", "reason", "recorded_at_utc"],
+        "properties": {
+          "backend": { "enum": ["mathematica", "primecount", "sagemath", "python"] },
+          "reason": { "enum": ["user_requested", "path_missing", "execution_failed", "version_mismatch"] },
+          "recorded_at_utc": { "type": "string", "format": "date-time" }
+        },
+        "additionalProperties": false
+      }
+    }
+  },
+  "additionalProperties": false
+}