pratyushpatodia-dev · Sep 2, 2026 · Sep 22, 2026 · Sep 22, 2026 · Sep 22, 2026 · Sep 22, 2026
diff --git a/data/fred/BAMLH0A0HYM2.parquet b/data/fred/BAMLH0A0HYM2.parquet
diff --git a/data/fred/CPIAUCSL.parquet b/data/fred/CPIAUCSL.parquet
diff --git a/data/fred/DFF.parquet b/data/fred/DFF.parquet
diff --git a/data/fred/DGS10.parquet b/data/fred/DGS10.parquet
diff --git a/data/fred/DGS2.parquet b/data/fred/DGS2.parquet
diff --git a/data/fred/FEDFUNDS.parquet b/data/fred/FEDFUNDS.parquet
diff --git a/data/fred/ICSA.parquet b/data/fred/ICSA.parquet
diff --git a/data/fred/IPMAN.parquet b/data/fred/IPMAN.parquet
diff --git a/data/fred/UNRATE.parquet b/data/fred/UNRATE.parquet
diff --git a/data/fred/VIXCLS.parquet b/data/fred/VIXCLS.parquet
diff --git a/data/yfinance/spy_adj_close_1d.parquet b/data/yfinance/spy_adj_close_1d.parquet
diff --git a/data/yfinance/xli_adj_close_1d.parquet b/data/yfinance/xli_adj_close_1d.parquet
diff --git a/implementations/__pycache__/__init__.cpython-312.pyc b/implementations/__pycache__/__init__.cpython-312.pyc
diff --git a/implementations/agentic_forecasting_implementations.egg-info/PKG-INFO b/implementations/agentic_forecasting_implementations.egg-info/PKG-INFO
@@ -0,0 +1,69 @@
+Metadata-Version: 2.4
+Name: agentic-forecasting-implementations
+Version: 0.1.0
+Summary: Reference method implementations for the Agentic Forecasting Bootcamp
+Author-email: Vector AI Engineering <ai_engineering@vectorinstitute.ai>
+Requires-Python: >=3.12
+Description-Content-Type: text/markdown
+Requires-Dist: aieng-forecasting[agentic,documents,llm,numerical]
+Requires-Dist: beautifulsoup4<5,>=4.12
+Requires-Dist: xgboost<4,>=2.1
+
+# implementations
+
+Self-contained reference implementations and their helper code.
+
+This is a local uv workspace package. It is installed automatically when you run `uv sync` from the repository root, but it is not a separately published public API.
+
+Some use cases are notebook-only. Others expose a small importable helper package so shared analysis, plotting, or data-registration code can live in Python modules instead of large notebook cells.
+
+---
+
+## Directory layout
+
+Numbered in the recommended order (mirrors the bootcamp progression: conventional numerical methods → LLM Processes → agents → agentic evaluation). The directories are not renamed — the numbers are an ordering convention used across the docs, and each directory stays an importable package (`from sp500_forecasting.data import ...`).
+
+```text
+implementations/
+|-- getting_started/          # 0 · CPI gasoline hello-world (start here)
+|   `-- specs/                #     backtest and eval YAML
+|-- sp500_forecasting/        # 1 · S&P 500 multivariate numerical comparison (financial markets)
+|   `-- specs/                #     backtest YAML (smoke + full)
+|-- food_price_forecasting/   # 2 · CFPR-style food CPI experiment
+|   `-- specs/                #     backtest YAML
+|-- energy_oil_forecasting/   # 3 · Daily WTI oil price forecasting experiment
+|   `-- specs/                #     backtest and eval YAML
+|-- boc_rate_decisions/       # 4 · Discrete-event reference: BoC cut/hold/hike direction
+|   `-- specs/                #     direction + binary backtest / eval / smoke YAML
+|-- tests/                    # tests for implementation-specific helper modules
+`-- pyproject.toml            # local workspace packaging
+```
+
+YAML backtest and eval specs live under each use case in `specs/`. Each directory is independent; see its `README.md` for the walkthrough. For the build-phase moves — onboarding data, standing up an experiment, customizing an agent, auditing a result — see [`guides/`](../guides/). To chat with the concierge or a domain starter in the ADK browser UI, see [`guides/05-access-adk-web-via-ssh-tunnel.md`](../guides/05-access-adk-web-via-ssh-tunnel.md) (includes the Coder SSH tunnel).
+
+Every domain use case (all except `getting_started`) also ships a `starter_agent/` module and a `99_starter_agent.ipynb` — a fresh, hackable **starter agent** that is the consistent "build your own" entry point for that use case (toggleable news search + code execution, two lightweight tool-usage skills, an interactive cell, and one scored forecast).
+
+`getting_started/` additionally ships a **`concierge_agent/`** module and **`99_repo_concierge.ipynb`** — a repo onboarding helper (not a forecaster) that answers questions about how the codebase works using a committed public-`main` knowledge digest. From the repository root: `uv run adk run implementations/getting_started/concierge_agent` (or `uv run adk web implementations/getting_started/concierge_agent` for the browser UI — [guide 5](../guides/05-access-adk-web-via-ssh-tunnel.md)). See [`getting_started/README.md`](getting_started/README.md) and the notebook for full usage.
+
+---
+
+## Relationship to `aieng-forecasting`
+
+- `aieng-forecasting` (`aieng.forecasting`) owns reusable infrastructure and reusable reference predictors under `aieng.forecasting.methods`.
+- `implementations/` owns use-case material: walkthrough notebooks, experiment-specific helper modules, plotting/analysis code, and task-specific framing.
+
+If code becomes broadly reusable across use cases, promote it into `aieng-forecasting`.
+
+---
+
+## Adding a new use case
+
+1. Create `implementations/<use-case>/`.
+2. Add a `README.md` describing the task, the data, and what the notebooks cover.
+3. Add YAML specs under `implementations/<use-case>/specs/`.
+4. Start with notebooks as the primary user surface.
+5. If notebook code becomes bulky or repeated, extract small helper modules into that use-case directory.
+6. Add tests under `implementations/tests/<use-case>/` for non-trivial helper logic.
+7. Promote code into `aieng-forecasting` once it is clearly reusable across more than one use case.
+
+For architecture principles and cross-cutting extension ideas, see `planning-docs/roadmap.md`.
diff --git a/implementations/agentic_forecasting_implementations.egg-info/SOURCES.txt b/implementations/agentic_forecasting_implementations.egg-info/SOURCES.txt
@@ -0,0 +1,76 @@
+README.md
+pyproject.toml
+agentic_forecasting_implementations.egg-info/PKG-INFO
+agentic_forecasting_implementations.egg-info/SOURCES.txt
+agentic_forecasting_implementations.egg-info/dependency_links.txt
+agentic_forecasting_implementations.egg-info/requires.txt
+agentic_forecasting_implementations.egg-info/top_level.txt
+boc_rate_decisions/__init__.py
+boc_rate_decisions/analysis.py
+boc_rate_decisions/data.py
+boc_rate_decisions/plots.py
+boc_rate_decisions/press_releases.py
+boc_rate_decisions/rationale_eval.py
+boc_rate_decisions/analyst_agent/__init__.py
+boc_rate_decisions/analyst_agent/agent.py
+boc_rate_decisions/predictors/__init__.py
+boc_rate_decisions/predictors/llmp_binary.py
+boc_rate_decisions/predictors/llmp_direction.py
+boc_rate_decisions/predictors/logistic_baseline.py
+boc_rate_decisions/starter_agent/__init__.py
+boc_rate_decisions/starter_agent/agent.py
+energy_oil_forecasting/__init__.py
+energy_oil_forecasting/analysis.py
+energy_oil_forecasting/data.py
+energy_oil_forecasting/paths.py
+energy_oil_forecasting/prophet_baseline.py
+energy_oil_forecasting/tasks.py
+energy_oil_forecasting/viz.py
+energy_oil_forecasting/adaptive_agent/__init__.py
+energy_oil_forecasting/adaptive_agent/agent.py
+energy_oil_forecasting/adaptive_agent/skill_state.py
+energy_oil_forecasting/adaptive_agent/skill_tools.py
+energy_oil_forecasting/adaptive_agent/curriculum/snapshot_utils.py
+energy_oil_forecasting/analyst_agent/__init__.py
+energy_oil_forecasting/analyst_agent/agent.py
+energy_oil_forecasting/starter_agent/__init__.py
+energy_oil_forecasting/starter_agent/agent.py
+energy_oil_forecasting/starter_agent/tools.py
+food_price_forecasting/__init__.py
+food_price_forecasting/analysis.py
+food_price_forecasting/data.py
+food_price_forecasting/plots.py
+food_price_forecasting/reports.py
+food_price_forecasting/smoke_report.py
+food_price_forecasting/predictors/__init__.py
+food_price_forecasting/predictors/llmp_quantile_grid.py
+food_price_forecasting/predictors/llmp_sampled_trajectory.py
+food_price_forecasting/starter_agent/__init__.py
+food_price_forecasting/starter_agent/agent.py
+getting_started/__init__.py
+getting_started/concierge_agent/__init__.py
+getting_started/concierge_agent/agent.py
+getting_started/concierge_agent/catalog.py
+getting_started/concierge_agent/catalog_build.py
+getting_started/concierge_agent/knowledge.py
+manufacturing_stress_forecasting/__init__.py
+manufacturing_stress_forecasting/data.py
+manufacturing_stress_forecasting/features.py
+manufacturing_stress_forecasting/run_agent_backtest.py
+manufacturing_stress_forecasting/run_agent_prediction.py
+manufacturing_stress_forecasting/run_smoke.py
+manufacturing_stress_forecasting/targets.py
+manufacturing_stress_forecasting/analyst_agent/__init__.py
+manufacturing_stress_forecasting/analyst_agent/agent.py
+manufacturing_stress_forecasting/predictors/__init__.py
+manufacturing_stress_forecasting/predictors/logistic.py
+manufacturing_stress_forecasting/predictors/xgboost.py
+sp500_forecasting/__init__.py
+sp500_forecasting/analysis.py
+sp500_forecasting/data.py
+sp500_forecasting/leaderboard.py
+sp500_forecasting/plots.py
+sp500_forecasting/predictors/__init__.py
+sp500_forecasting/predictors/llmp_sampled_trajectory.py
+sp500_forecasting/starter_agent/__init__.py
+sp500_forecasting/starter_agent/agent.py
diff --git a/implementations/agentic_forecasting_implementations.egg-info/dependency_links.txt b/implementations/agentic_forecasting_implementations.egg-info/dependency_links.txt
@@ -0,0 +1 @@
+
diff --git a/implementations/agentic_forecasting_implementations.egg-info/requires.txt b/implementations/agentic_forecasting_implementations.egg-info/requires.txt
@@ -0,0 +1,3 @@
+aieng-forecasting[agentic,documents,llm,numerical]
+beautifulsoup4<5,>=4.12
+xgboost<4,>=2.1
diff --git a/implementations/agentic_forecasting_implementations.egg-info/top_level.txt b/implementations/agentic_forecasting_implementations.egg-info/top_level.txt
@@ -0,0 +1,6 @@
+boc_rate_decisions
+energy_oil_forecasting
+food_price_forecasting
+getting_started
+manufacturing_stress_forecasting
+sp500_forecasting
diff --git a/implementations/manufacturing_stress_forecasting/README.md b/implementations/manufacturing_stress_forecasting/README.md
@@ -0,0 +1,118 @@
+# Manufacturing stress forecasting
+
+This implementation asks one Track 1 question:
+
+> Given information available at a monthly forecast origin, what is the
+> probability that U.S. manufacturing will be under stress three months later?
+
+The feature service provides trailing 1-, 3-, and 6-month IPMAN changes plus
+the requested FRED and Yahoo Finance fields: `FEDFUNDS`, `YC_SPREAD`,
+`CPIAUCSL`, `CPI_YOY`, `UNRATE`, `ICSA`, `VIXCLS`, `HY_SPREAD`, and 3- and
+12-month returns for both `SPY` and `XLI`. FRED levels are collapsed to
+monthly observations, derived fields use the documented source series, and
+Yahoo returns use monthly adjusted-close prices.
+
+## Target
+
+A month is labelled `1` (stress) when IPMAN has declined by at least 2% over
+its preceding three months; otherwise it is `0`. The threshold is a provisional
+version-1 definition and should be reviewed visually before expanding the
+project.
+
+The forecast made at month `t` predicts the stress label at `t + 3 months`.
+That distinction makes this forecasting rather than current-state detection.
+
+## Predictors
+
+- `HistoricalFrequencyPredictor`: the visible historical stress rate.
+- `ManufacturingStressLogisticPredictor`: fit-at-origin logistic regression on
+  the IPMAN and macro variables.
+- `ManufacturingStressXGBoostPredictor`: a small fit-at-origin gradient-boosted
+  tree classifier using the same IPMAN and macro variables and cutoff-safe
+  training rows.
+- `manufacturing_stress_analyst`: a structured LLM predictor receiving the same
+  cutoff-safe IPMAN and macro signals plus recent IPMAN history and historical
+  base rates.
+
+All predictors return `BinaryForecast` probabilities; backtested predictors are scored with Brier score.
+
+## Data and cutoff assumptions
+Compare XGBoost with logistic regression and historical frequency rather than judging it
+in isolation, because this small monthly dataset can overfit flexible models.
+`FREDAdapter` caches the required FRED series under `data/fred/`, and
+`YFinanceDailyAdapter` caches `SPY` and `XLI` under `data/yfinance/`.
+Yahoo refreshes request history from 1998 onward explicitly so the provider's
+default recent-history window cannot replace the long-term cache.
+IPMAN is conservatively treated as available one month after its reference
+month. Daily rate observations are treated as available on the next business
+day and collapsed to their final monthly observation. The standard FRED API
+does not provide full point-in-time vintages, so historical observations may
+still contain later revisions; a production study should use ALFRED vintages.
+
+## Run
+
+The primary interactive entry point is
+[`manufacturing_stress_workbench.ipynb`](manufacturing_stress_workbench.ipynb).
+Open it in VS Code or Jupyter and use its configuration cell to refresh FRED
+data, run the deterministic smoke test, and explicitly opt in to the cached
+LLMP backtest without using the terminal.
+
+From the repository root, put a personal FRED key in `.env` or export it:
+
+```bash
+export FRED_API_KEY="..."
+```
+
+Populate the cache and inspect the registered series:
+
+```bash
+uv run python scripts/fetch_manufacturing_stress.py
+```
+
+Run the deterministic small backtest:
+
+```bash
+uv run --directory implementations python -m manufacturing_stress_forecasting.run_smoke
+```
+
+The output prints one mean Brier score per predictor; lower is better. The
+logistic model should be compared against historical frequency, not judged in
+isolation.
+
+Run the token-limited LLMP backtest explicitly:
+
+```bash
+uv run --directory implementations python -m manufacturing_stress_forecasting.run_agent_backtest
+```
+
+This evaluates historical frequency, logistic regression, XGBoost, and the
+`manufacturing_stress_analyst` agent through the same binary backtest and
+Brier-score calculation. The agent run uses the default lite model, a
+12-month IPMAN history, compact JSON prompts, a 384-token response cap, and
+one retry per failed origin. Calendar dates are replaced by relative month
+offsets in retrospective agent prompts to reduce historical-event recall.
+
+Complete results are cached under a specification-fingerprinted directory in
+`data/predictions/`, so changing the stride, horizon, dates, or warmup cannot
+silently reuse an incompatible result. Incomplete runs with skipped origins
+are not cached. Use `--force-refresh` to intentionally re-run every predictor.
+The command also verifies that every reported model was scored on the exact
+same origin and forecast-date pairs.
+
+This remains a retrospective LLM pseudo-backtest: anonymizing dates reduces,
+but cannot eliminate, the possibility that a modern model recognizes a
+historical episode from its training knowledge. Use prospectively recorded
+forecasts for a clean out-of-sample LLM evaluation.
+
+Run one current forecast, including the structured agent:
+
+```bash
+uv run --directory implementations python -m manufacturing_stress_forecasting.run_agent_prediction
+```
+
+## Next steps
+
+1. Plot IPMAN and the derived stress months; confirm or revise the 2% threshold.
+2. Compare the expanded macro-panel score with the earlier IPMAN-only result.
+3. Compare the cached agent backtest against the deterministic baselines only
+  after checking scored and skipped origin counts.
diff --git a/implementations/manufacturing_stress_forecasting/__init__.py b/implementations/manufacturing_stress_forecasting/__init__.py
@@ -0,0 +1,16 @@
+"""Minimal IPMAN-based manufacturing-stress forecasting use case."""
+
+from manufacturing_stress_forecasting.data import (
+    IPMAN_SERIES_ID,
+    STRESS_SERIES_ID,
+    build_manufacturing_stress_service,
+)
+from manufacturing_stress_forecasting.predictors import ManufacturingStressLogisticPredictor
+
+
+__all__ = [
+    "IPMAN_SERIES_ID",
+    "STRESS_SERIES_ID",
+    "ManufacturingStressLogisticPredictor",
+    "build_manufacturing_stress_service",
+]
diff --git a/implementations/manufacturing_stress_forecasting/__pycache__/__init__.cpython-312.pyc b/implementations/manufacturing_stress_forecasting/__pycache__/__init__.cpython-312.pyc
diff --git a/implementations/manufacturing_stress_forecasting/__pycache__/data.cpython-312.pyc b/implementations/manufacturing_stress_forecasting/__pycache__/data.cpython-312.pyc
diff --git a/implementations/manufacturing_stress_forecasting/__pycache__/features.cpython-312.pyc b/implementations/manufacturing_stress_forecasting/__pycache__/features.cpython-312.pyc
diff --git a/...entations/manufacturing_stress_forecasting/__pycache__/run_agent_backtest.cpython-312.pyc b/...entations/manufacturing_stress_forecasting/__pycache__/run_agent_backtest.cpython-312.pyc
diff --git a/...tations/manufacturing_stress_forecasting/__pycache__/run_agent_prediction.cpython-312.pyc b/...tations/manufacturing_stress_forecasting/__pycache__/run_agent_prediction.cpython-312.pyc
diff --git a/implementations/manufacturing_stress_forecasting/__pycache__/run_smoke.cpython-312.pyc b/implementations/manufacturing_stress_forecasting/__pycache__/run_smoke.cpython-312.pyc
diff --git a/implementations/manufacturing_stress_forecasting/__pycache__/targets.cpython-312.pyc b/implementations/manufacturing_stress_forecasting/__pycache__/targets.cpython-312.pyc
diff --git a/implementations/manufacturing_stress_forecasting/analyst_agent/__init__.py b/implementations/manufacturing_stress_forecasting/analyst_agent/__init__.py
@@ -0,0 +1,14 @@
+"""Quantitative-only manufacturing-stress analyst agent."""
+
+from manufacturing_stress_forecasting.analyst_agent.agent import (
+    ManufacturingStressPromptBuilder,
+    build_manufacturing_stress_agent_config,
+    build_manufacturing_stress_agent_predictor,
+)
+
+
+__all__ = [
+    "ManufacturingStressPromptBuilder",
+    "build_manufacturing_stress_agent_config",
+    "build_manufacturing_stress_agent_predictor",
+]
diff --git a/...tions/manufacturing_stress_forecasting/analyst_agent/__pycache__/__init__.cpython-312.pyc b/...tions/manufacturing_stress_forecasting/analyst_agent/__pycache__/__init__.cpython-312.pyc
diff --git a/...ntations/manufacturing_stress_forecasting/analyst_agent/__pycache__/agent.cpython-312.pyc b/...ntations/manufacturing_stress_forecasting/analyst_agent/__pycache__/agent.cpython-312.pyc