HF README: fix CLI flags (plural), document mock model, lb90 result paths

README.md

@@ -46,7 +46,7 @@ dataset_info:
 
 # TeamBench
 
-A multi-agent coordination benchmark with **OS-enforced** Planner / Executor / Verifier role separation. **931 evaluation instances · 19 categories · 5 ablation conditions.**
+A benchmark for evaluating multi-agent LLM coordination under **OS-enforced** Planner / Executor / Verifier role separation. **931 evaluation instances · 19 categories · 5 ablation conditions.**
 
 This Hugging Face dataset is the structured metadata layer. The task workspaces, generators, deterministic graders, and the evaluation harness live in the GitHub repository: <https://github.com/ybkim95/TeamBench>.
 
@@ -78,7 +78,8 @@ The dataset metadata is enough to filter and inspect tasks. To actually score a
 ```bash
 git clone https://github.com/ybkim95/TeamBench.git
 cd TeamBench
-pip install -e .
+pip install -e ".[all]"
+docker compose build
 
 export ANTHROPIC_API_KEY=...
 export OPENAI_API_KEY=...
@@ -86,13 +87,22 @@ export GEMINI_API_KEY=...
 
 # Run a task across all 5 ablation conditions, seed 0
 python -m harness.ablation \
-    --task DIST1_queue_race \
     --model gemini-3-flash-preview \
-    --seed 0 \
-    --out shared/runs/example
+    --tasks DIST1_queue_race \
+    --seeds 0 \
+    --conditions oracle restricted full team_no_plan team_no_verify \
+    --output shared/runs/example
 ```
 
-Output: `shared/runs/example/<task>/<condition>/score.json` with `passed` and a partial score in `[0,1]`. Aggregate across tasks with `python -m harness.compute_tni --runs-dir shared/runs/example`.
+The deterministic grader emits `score.json` per (task, condition) under `shared/runs/example/`. Aggregate with:
+
+```bash
+python -m harness.compute_tni \
+    --ablation shared/ablation_results/lb90_<your-model>_seed0.json \
+    --output shared/ablation_results/tni_<your-model>.json
+```
+
+No keys handy? `--model mock` runs an in-process stub that exercises the grader pipeline end-to-end without provider calls.
 
 ## Five conditions