claude-turing 4.7.0 → 4.8.1

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "turing",
-  "version": "4.7.0",
-  "description": "Autonomous ML research harness — the autoresearch loop as a formal protocol. 74 commands, 2 specialized agents, modern skills/turing package mirror, operational intelligence (postmortem + doctor + plan), model lifecycle (update + registry), what-if analysis (whatif + counterfactual + simulate), collaboration (onboard + share + review), research communication (cite + present + changelog), experiment archaeology (trend + flashback + archive + annotate + search + template + replay), model surgery (prune + quantize + merge + surgery), feature & training intelligence, model debugging, pre-training intelligence, meta-intelligence, scaling & efficiency, model composition, deep analysis, experiment orchestration, literature + paper, model export, profiling, checkpoints, experiment intelligence, statistical rigor, tree-search, cost-performance, model cards, hypothesis database, novelty guard, anti-cheating, taste-leverage loop. Inspired by Karpathy's autoresearch and the scientific method itself.",
+  "version": "4.8.1",
+  "description": "Autonomous ML research harness — the autoresearch loop as a formal protocol. 74 commands, 2 specialized agents, skills/turing source layout, operational intelligence (postmortem + doctor + plan), model lifecycle (update + registry), what-if analysis (whatif + counterfactual + simulate), collaboration (onboard + share + review), research communication (cite + present + changelog), experiment archaeology (trend + flashback + archive + annotate + search + template + replay), model surgery (prune + quantize + merge + surgery), feature & training intelligence, model debugging, pre-training intelligence, meta-intelligence, scaling & efficiency, model composition, deep analysis, experiment orchestration, literature + paper, model export, profiling, checkpoints, experiment intelligence, statistical rigor, tree-search, cost-performance, model cards, hypothesis database, novelty guard, anti-cheating, taste-leverage loop. Inspired by Karpathy's autoresearch and the scientific method itself.",
   "author": {
     "name": "Prannaya Gupta"
   },

package/README.md

@@ -3,7 +3,7 @@
 *The research assistant that can't fool itself.*
 
 <p align="center">
-  <img src="https://img.shields.io/badge/version-4.7.0-ffb74d?style=flat-square&labelColor=1a1a2e" alt="Version" /> 
+  <img src="https://img.shields.io/badge/version-4.8.1-ffb74d?style=flat-square&labelColor=1a1a2e" alt="Version" />
   <img src="https://img.shields.io/badge/license-MIT-ff4d4d?style=flat-square&labelColor=1a1a2e" alt="License" /> 
   <img src="https://img.shields.io/badge/Claude_Code-plugin-ff4d4d?style=flat-square&labelColor=1a1a2e" alt="Claude Code" /> 
   <img src="https://img.shields.io/badge/Node.js-20%2B-ff4d4d?style=flat-square&labelColor=1a1a2e" alt="Node.js" />

package/agents/ml-evaluator.md

@@ -22,13 +22,13 @@ In quantum mechanics, observation changes the system. In ML experimentation, the
 
 ## Useful Commands
 
-Always activate the venv first: `source .venv/bin/activate`
+Always run Python through uv from the ML directory.
 
 | Command | Purpose |
 |---------|---------|
-| `python scripts/show_metrics.py --last 10` | Recent experiment summary |
-| `python scripts/compare_runs.py <a> <b>` | Side-by-side comparison |
-| `python evaluate.py` | Run evaluation on current model |
+| `uv run python scripts/show_metrics.py --last 10` | Recent experiment summary |
+| `uv run python scripts/compare_runs.py <a> <b>` | Side-by-side comparison |
+| `uv run python evaluate.py` | Run evaluation on current model |
 | `cat experiments/results.tsv` | Quick-reference TSV |
 
 ## Analysis Framework

package/agents/ml-researcher.md

@@ -27,8 +27,8 @@ Read `program.md` in the ML directory for the complete experiment loop protocol.
 ## Constraints
 
 - **Only modify `train.py` and `config.yaml`.** `evaluate.py` is HIDDEN (do not read or reference). Other pipeline files are READ-ONLY.
-- **Always work in the venv:** `source .venv/bin/activate`
-- **Redirect training output:** `python train.py > run.log 2>&1`
+- **Always run Python through uv:** `uv run python ...`
+- **Redirect training output:** `uv run python train.py > run.log 2>&1`
 - **Parse metrics with grep:** `grep -A 10 "^---" run.log | head -10`
 - **Use @ml-evaluator** for analysis tasks — it has no Write/Edit tools and cannot accidentally break the pipeline.
 

package/bin/turing-init.sh

@@ -33,8 +33,8 @@ echo ""
 if [[ $# -eq 0 ]] || [[ "${1:-}" == "--interactive" ]]; then
     python3 "$SCAFFOLD_SCRIPT" --interactive --templates-dir "$TEMPLATES_DIR" --no-venv
     echo ""
-    echo "  To set up the virtual environment:"
-    echo "    cd <ml_dir> && python3 -m venv .venv && source .venv/bin/activate && pip install -r requirements.txt"
+    echo "  To set up the uv environment:"
+    echo "    cd <ml_dir> && uv sync"
     exit 0
 fi
 

package/commands/ablate.md

@@ -1,7 +1,6 @@
 ---
 name: ablate
 description: Run systematic ablation study — remove components one at a time, measure impact, produce publication-ready table with dead-weight flagging.
-disable-model-invocation: true
 argument-hint: "[exp-id] [--components \"X,Y\"] [--seeds 3] [--latex]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Run a systematic ablation study to measure the contribution of each model compon
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -23,7 +22,7 @@ Run a systematic ablation study to measure the contribution of each model compon
 
 3. **Run ablation study:**
    ```bash
-   python scripts/ablation_study.py $ARGUMENTS
+   uv run python scripts/ablation_study.py $ARGUMENTS
    ```
 
 4. **Report results:**

package/commands/annotate.md

@@ -1,7 +1,6 @@
 ---
 name: annotate
 description: Retrospective experiment annotations — add human notes, tags, and context that automated metrics can't capture.
-disable-model-invocation: true
 argument-hint: "<exp-id> \"note\" [--tag fragile] | --list | --search \"keyword\""
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 Add context that experiment logs can't capture. "This only worked because the data was pre-sorted."
 
 ## Steps
-1. **Activate environment:** `source .venv/bin/activate`
-2. **Run:** `python scripts/experiment_annotations.py $ARGUMENTS`
+1. **Sync environment:** `uv sync`
+2. **Run:** `uv run python scripts/experiment_annotations.py $ARGUMENTS`
 3. **Operations:** add (text + tags), list (per-experiment or all), search (keyword or tag)
 4. **Stored in:** `experiments/annotations.yaml`
 

package/commands/archive.md

@@ -1,7 +1,6 @@
 ---
 name: archive
 description: Experiment lifecycle cleanup — compress old artifacts, prune checkpoints, create queryable summary index. Reclaim disk space.
-disable-model-invocation: true
 argument-hint: "[--older-than 30d] [--keep-best 10] [--dry-run]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 Keep your project directory manageable after 200+ experiments.
 
 ## Steps
-1. **Activate environment:** `source .venv/bin/activate`
-2. **Run:** `python scripts/experiment_archive.py $ARGUMENTS`
+1. **Sync environment:** `uv sync`
+2. **Run:** `uv run python scripts/experiment_archive.py $ARGUMENTS`
 3. **Protected experiments:** Pareto-optimal, current best, recent, top-N by metric
 4. **Report:** archived count, preserved count, space reclaimed
 5. **Saved output:** `experiments/archive/index.yaml`

package/commands/audit.md

@@ -1,7 +1,6 @@
 ---
 name: audit
 description: Pre-submission methodology audit — catch data leakage, missing baselines, cherry-picked seeds, and incomplete ablations before a reviewer does.
-disable-model-invocation: true
 argument-hint: "[--strict] [--checklist neurips]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ A reviewer checklist you run before submitting. Catches methodology mistakes tha
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -22,7 +21,7 @@ A reviewer checklist you run before submitting. Catches methodology mistakes tha
 
 3. **Run methodology audit:**
    ```bash
-   python scripts/methodology_audit.py $ARGUMENTS
+   uv run python scripts/methodology_audit.py $ARGUMENTS
    ```
 
 4. **Checks performed:**

package/commands/baseline.md

@@ -1,7 +1,6 @@
 ---
 name: baseline
 description: Automatic baseline generation — random, majority/mean, linear, k-NN baselines in 60 seconds. Every experiment needs a "is this better than dumb?" reference.
-disable-model-invocation: true
 argument-hint: "[--methods all|simple|linear] [--data data.npz]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Generate trivial baselines so you always know if your model is meaningfully bett
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -22,7 +21,7 @@ Generate trivial baselines so you always know if your model is meaningfully bett
 
 3. **Run baseline generation:**
    ```bash
-   python scripts/generate_baselines.py $ARGUMENTS
+   uv run python scripts/generate_baselines.py $ARGUMENTS
    ```
 
 4. **Baselines generated:**

package/commands/brief.md

@@ -1,9 +1,8 @@
 ---
 name: brief
 description: Generate a structured research intelligence report from experiment history — what's been learned, what's promising, what's exhausted, and what the human should consider next. Use --deep for literature-grounded suggestions.
-disable-model-invocation: true
 argument-hint: "[ml/project] [--deep]"
-allowed-tools: Read, Bash(python scripts/*:*, source .venv/bin/activate:*), Grep, Glob, WebSearch, WebFetch
+allowed-tools: Read, Bash(uv run python scripts/*:*, uv sync:*), Grep, Glob, WebSearch, WebFetch
 ---
 
 Generate a research briefing that a human can read in 2 minutes and immediately decide what to inject next.
@@ -24,14 +23,14 @@ Before generating the briefing, detect which project to report on:
 
 1. **Generate the briefing:**
    ```bash
-   source .venv/bin/activate && python scripts/generate_brief.py
+   uv run python scripts/generate_brief.py
    ```
 
 2. **Self-critique the briefing** before presenting. Review the generated output and check:
    - **Recommendations specificity:** Are they concrete enough to act on? "Try a different model" is bad. "Try LightGBM with leaf-wise growth because exp-004 showed depth sensitivity" is good. If vague, rewrite them with specific model/hyperparameter suggestions grounded in the experiment data.
    - **Exhausted directions coverage:** Cross-reference the "Model Types Explored" section against `experiments/log.jsonl`. Are there discarded experiments missing from the summary? If so, add them.
    - **Convergence estimate grounding:** If the briefing says "close to convergence" or "further improvement possible", verify against the actual metric trajectory. Is the claim supported by the numbers?
-   - **Metric accuracy:** Spot-check that the "Current Best" metrics match the actual log. Run `python scripts/show_metrics.py --last 1` if uncertain.
+   - **Metric accuracy:** Spot-check that the "Current Best" metrics match the actual log. Run `uv run python scripts/show_metrics.py --last 1` if uncertain.
 
    If any section fails the check, regenerate just that section. Max 1 revision round — don't over-polish.
 
@@ -76,7 +75,7 @@ When `--deep` is requested, add a 7th section: **Literature-Grounded Suggestions
 
 4. **Queue suggestions** as hypotheses:
    ```bash
-   source .venv/bin/activate && python scripts/manage_hypotheses.py add "<technique>: <rationale> (source: <citation>)" --priority medium --source literature
+   uv run python scripts/manage_hypotheses.py add "<technique>: <rationale> (source: <citation>)" --priority medium --source literature
    ```
 
 5. **Format as a section** appended to the briefing.
@@ -84,7 +83,7 @@ When `--deep` is requested, add a 7th section: **Literature-Grounded Suggestions
 ## Saving Briefs
 
 ```bash
-mkdir -p briefs && python scripts/generate_brief.py > briefs/brief-$(date +%Y-%m-%d).md
+mkdir -p briefs && uv run python scripts/generate_brief.py > briefs/brief-$(date +%Y-%m-%d).md
 ```
 
 ## When to Use

package/commands/budget.md

@@ -1,7 +1,6 @@
 ---
 name: budget
 description: Compute budget manager — set experiment/time limits, track allocation across explore/exploit phases, auto-shift modes, hard stop.
-disable-model-invocation: true
 argument-hint: "<set|status|reset> [--experiments 50] [--hours 8]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Set a compute ceiling and let the system optimize within it. Prevents runaway ex
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -23,7 +22,7 @@ Set a compute ceiling and let the system optimize within it. Prevents runaway ex
 
 3. **Run budget manager:**
    ```bash
-   python scripts/budget_manager.py $ARGUMENTS
+   uv run python scripts/budget_manager.py $ARGUMENTS
    ```
 
 4. **Actions:**

package/commands/calibrate.md

@@ -1,7 +1,6 @@
 ---
 name: calibrate
 description: Probability calibration — measure ECE, plot reliability diagrams, apply Platt scaling or isotonic regression.
-disable-model-invocation: true
 argument-hint: "[exp-id] [--method platt|isotonic|temperature|auto]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Make model probabilities trustworthy. Does 80% confidence actually mean 80% corr
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -22,7 +21,7 @@ Make model probabilities trustworthy. Does 80% confidence actually mean 80% corr
 
 3. **Run calibration:**
    ```bash
-   python scripts/calibration.py $ARGUMENTS
+   uv run python scripts/calibration.py $ARGUMENTS
    ```
 
 4. **Report includes:**

package/commands/card.md

@@ -1,8 +1,7 @@
 ---
 name: card
 description: Generate a standardized model card documenting the trained model — type, performance, training data, limitations, intended use, and artifact contract.
-disable-model-invocation: true
-allowed-tools: Read, Bash(python scripts/*:*, source .venv/bin/activate:*), Grep, Glob
+allowed-tools: Read, Bash(uv run python scripts/*:*, uv sync:*), Grep, Glob
 ---
 
 You generate a standardized model card from the experiment log, model contract, and config.
@@ -11,12 +10,12 @@ You generate a standardized model card from the experiment log, model contract,
 
 1. **Activate the virtual environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Run the model card generator:**
    ```bash
-   python scripts/generate_model_card.py --config config.yaml --log experiments/log.jsonl --contract model_contract.md --output MODEL_CARD.md
+   uv run python scripts/generate_model_card.py --config config.yaml --log experiments/log.jsonl --contract model_contract.md --output MODEL_CARD.md
    ```
 
 3. **Read and present the generated card:**

package/commands/changelog.md

@@ -1,7 +1,6 @@
 ---
 name: changelog
 description: Model changelog generation — auto-generate human-readable progress narrative from experiment history for stakeholders.
-disable-model-invocation: true
 argument-hint: "[--since exp-id|date] [--audience technical|stakeholder]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 Translate experiment logs into a narrative that PMs and stakeholders can read in 2 minutes.
 
 ## Steps
-1. **Activate environment:** `source .venv/bin/activate`
-2. **Run:** `python scripts/generate_changelog.py $ARGUMENTS`
+1. **Sync environment:** `uv sync`
+2. **Run:** `uv run python scripts/generate_changelog.py $ARGUMENTS`
 3. **Audience:** technical (experiment IDs, configs), stakeholder (plain English, percentages)
 4. **Saved output:** `paper/CHANGELOG.md`
 

package/commands/checkpoint.md

@@ -1,7 +1,6 @@
 ---
 name: checkpoint
 description: Smart checkpoint management — list, prune (Pareto-based), average top-K, resume from any point, disk usage stats.
-disable-model-invocation: true
 argument-hint: "<list|prune|average|resume|stats> [exp-id] [--top 3] [--dry-run]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Manage model checkpoints intelligently using Pareto dominance.
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -23,7 +22,7 @@ Manage model checkpoints intelligently using Pareto dominance.
 
 3. **Run checkpoint manager:**
    ```bash
-   python scripts/checkpoint_manager.py $ARGUMENTS
+   uv run python scripts/checkpoint_manager.py $ARGUMENTS
    ```
 
 4. **Report results by action:**

package/commands/cite.md

@@ -1,7 +1,6 @@
 ---
 name: cite
 description: Citation & attribution manager — track papers, datasets, methods. Audit for missing citations, generate BibTeX.
-disable-model-invocation: true
 argument-hint: "<add|list|check|bib> [--key Chen2016 --title XGBoost --url ...]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 Track which papers and methods influenced each experiment. Catch missing citations before submission.
 
 ## Steps
-1. **Activate environment:** `source .venv/bin/activate`
-2. **Run:** `python scripts/citation_manager.py $ARGUMENTS`
+1. **Sync environment:** `uv sync`
+2. **Run:** `uv run python scripts/citation_manager.py $ARGUMENTS`
 3. **Operations:** add (associate citation with experiment), list (group by type), check (audit missing), bib (BibTeX)
 4. **Stored in:** `experiments/citations.yaml`
 

package/commands/compare.md

@@ -1,7 +1,6 @@
 ---
 name: compare
 description: Compare two ML experiment runs side-by-side — metrics, configuration deltas, and a verdict on which approach is more promising.
-disable-model-invocation: true
 argument-hint: "<exp-id-1> <exp-id-2>"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -12,7 +11,7 @@ Compare two ML experiment runs side-by-side to understand what changed and why o
 
 1. **Run comparison:**
    ```bash
-   source .venv/bin/activate && python scripts/compare_runs.py $0 $1
+   uv run python scripts/compare_runs.py $0 $1
    ```
 
 2. **Analyze the delta:**

package/commands/counterfactual.md

@@ -1,7 +1,6 @@
 ---
 name: counterfactual
 description: Input-level counterfactual explanations — find the smallest input change to flip a prediction.
-disable-model-invocation: true
 argument-hint: "<exp-id> --sample <index> [--target <class>]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 What would need to change to flip this prediction? Minimum-change counterfactual for individual predictions.
 
 ## Steps
-1. `source .venv/bin/activate`
-2. `python scripts/counterfactual_explanation.py $ARGUMENTS`
+1. `uv sync`
+2. `uv run python scripts/counterfactual_explanation.py $ARGUMENTS`
 3. **Saved:** `experiments/counterfactuals/`
 
 ## Methods

package/commands/curriculum.md

@@ -1,7 +1,6 @@
 ---
 name: curriculum
 description: Training curriculum optimization — order data by difficulty, compare easy-to-hard vs hard-to-easy vs self-paced strategies.
-disable-model-invocation: true
 argument-hint: "[exp-id] [--strategies easy-to-hard,random]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Does the order your model sees data matter? Find out systematically.
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -22,7 +21,7 @@ Does the order your model sees data matter? Find out systematically.
 
 3. **Run curriculum analysis:**
    ```bash
-   python scripts/curriculum_optimizer.py $ARGUMENTS
+   uv run python scripts/curriculum_optimizer.py $ARGUMENTS
    ```
 
 4. **Strategies tested:**

package/commands/design.md

@@ -1,9 +1,8 @@
 ---
 name: design
 description: Generate a structured experiment design for a hypothesis. Reads experiment history, searches literature for methodology, produces a scored design document at experiments/designs/.
-disable-model-invocation: true
 argument-hint: "<hypothesis-id or description>"
-allowed-tools: Read, Write, Bash(python scripts/*:*, source .venv/bin/activate:*, mkdir:*), Grep, Glob, WebSearch, WebFetch
+allowed-tools: Read, Write, Bash(uv run python scripts/*:*, uv sync:*, mkdir:*), Grep, Glob, WebSearch, WebFetch
 ---
 
 Front-load the thinking before the coding. Given a hypothesis, produce a structured experiment design grounded in methodology from the literature.
@@ -14,7 +13,7 @@ Front-load the thinking before the coding. Given a hypothesis, produce a structu
 
 If `$ARGUMENTS` matches `hyp-NNN`, load the hypothesis:
 ```bash
-source .venv/bin/activate && python scripts/manage_hypotheses.py show $ARGUMENTS
+uv run python scripts/manage_hypotheses.py show $ARGUMENTS
 ```
 
 If freeform text, use it directly as the hypothesis description.
@@ -24,7 +23,7 @@ Read the current config and experiment state:
 cat config.yaml
 ```
 ```bash
-source .venv/bin/activate && python scripts/show_metrics.py --last 10 2>/dev/null || echo "No experiments yet"
+uv run python scripts/show_metrics.py --last 10 2>/dev/null || echo "No experiments yet"
 ```
 ```bash
 cat experiment_state.yaml 2>/dev/null || echo "No experiment state yet"

package/commands/diagnose.md

@@ -1,7 +1,6 @@
 ---
 name: diagnose
 description: Error analysis — cluster failure cases, identify systematic failure modes, and suggest targeted fixes with auto-queued hypotheses.
-disable-model-invocation: true
 argument-hint: "[exp-id] [--auto-queue] [--top 5]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,15 +9,15 @@ Analyze where and why the model fails, beyond aggregate metrics.
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Generate predictions if needed:**
    Check if `experiments/predictions/exp-NNN-preds.yaml` exists. If not, run:
    ```bash
-   python train.py --predict-only --output experiments/predictions/
+   uv run python train.py --predict-only --output experiments/predictions/
    ```
    The predictions file must contain `y_true`, `y_pred`, `task_type`, and optionally `features`.
 
@@ -29,7 +28,7 @@ Analyze where and why the model fails, beyond aggregate metrics.
 
 4. **Run error analysis:**
    ```bash
-   python scripts/diagnose_errors.py $ARGUMENTS
+   uv run python scripts/diagnose_errors.py $ARGUMENTS
    ```
 
 5. **Report results:**

package/commands/diff.md

@@ -1,7 +1,6 @@
 ---
 name: diff
 description: Deep experiment comparison — config diffs, metric significance, per-class regressions, training curve divergence, feature importance shifts.
-disable-model-invocation: true
 argument-hint: "<exp-a> <exp-b> [--code]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Deep diagnostic comparison of two experiments. Goes beyond "which metric is high
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -22,7 +21,7 @@ Deep diagnostic comparison of two experiments. Goes beyond "which metric is high
 
 3. **Run deep comparison:**
    ```bash
-   python scripts/experiment_diff.py $ARGUMENTS
+   uv run python scripts/experiment_diff.py $ARGUMENTS
    ```
 
 4. **Report results — the diff includes:**

package/commands/distill.md

@@ -1,7 +1,6 @@
 ---
 name: distill
 description: Model compression via distillation — train a smaller student model to match a larger teacher's predictions.
-disable-model-invocation: true
 argument-hint: "<teacher-exp-id> [--compression 4] [--method soft-labels]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Compress a large model into a smaller, faster one for production. Measures the a
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -24,7 +23,7 @@ Compress a large model into a smaller, faster one for production. Measures the a
 
 3. **Run distillation planner:**
    ```bash
-   python scripts/model_distiller.py $ARGUMENTS
+   uv run python scripts/model_distiller.py $ARGUMENTS
    ```
 
 4. **Report includes:**

package/commands/doctor.md

@@ -1,7 +1,6 @@
 ---
 name: doctor
 description: Harness self-diagnosis — check environment, project, resources, and git state. Auto-fix common issues.
-disable-model-invocation: true
 argument-hint: "[--fix] [--verbose]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -9,8 +8,8 @@ allowed-tools: Read, Bash(*), Grep, Glob
 Is Turing healthy? Check everything and get a score.
 
 ## Steps
-1. `source .venv/bin/activate`
-2. `python scripts/harness_doctor.py $ARGUMENTS`
+1. `uv sync`
+2. `uv run python scripts/harness_doctor.py $ARGUMENTS`
 3. **Saved:** `experiments/doctor/`
 
 ## Checks

package/commands/ensemble.md

@@ -1,7 +1,6 @@
 ---
 name: ensemble
 description: Automated ensemble construction — combines top-K models via voting, stacking, and blending for zero-cost improvement.
-disable-model-invocation: true
 argument-hint: "[--top-k 5] [--methods voting,stacking,blending]"
 allowed-tools: Read, Bash(*), Grep, Glob
 ---
@@ -10,9 +9,9 @@ Build ensembles from your best experiments automatically. Often yields 1-3% impr
 
 ## Steps
 
-1. **Activate environment:**
+1. **Sync environment:**
    ```bash
-   source .venv/bin/activate
+   uv sync
    ```
 
 2. **Parse arguments from `$ARGUMENTS`:**
@@ -23,7 +22,7 @@ Build ensembles from your best experiments automatically. Often yields 1-3% impr
 
 3. **Run ensemble construction:**
    ```bash
-   python scripts/build_ensemble.py $ARGUMENTS
+   uv run python scripts/build_ensemble.py $ARGUMENTS
    ```
 
 4. **Report results:**