# Robotics / Embodied AI Aiwiki — Schema

## Deep-loop sweep JSON output

Each sweep agent returns an array of paper objects with this exact frontmatter:

```json
[
  {
    "paper_id": "arxiv:2412.01234" | "openreview:abc123" | "doi:10.xxxx/...",
    "title": "...",
    "authors": ["First Last", ...],
    "date": "YYYY-MM",
    "venue": "RSS 2024" | "CoRL 2024" | "ICRA 2025" | "Nature 2024" | "arxiv:cs.RO 2024-12" | ...,
    "summary": "<3-sentence summary; focus on what closure mechanism the claim engages>",
    "candidate_bill": "Bill_1" | ... | "Bill_13" | null,
    "candidate_meta_cost": "M1" | ... | "M6" | null,
    "verdict": "known_bill" | "needs_gate" | "out_of_scope" | "needs_gate_declaration" | "rebuttal_paper",
    "confidence": 0.0..1.0,
    "watchlist_tier": "monthly" | "quarterly" | "triggered" | null,
    "system_class": "VLA" | "manipulation_specialist" | "locomotion_specialist" | "navigation" | "driving" | "humanoid" | "tele-operation_baseline" | "other:<name>",
    "embodiment": "Google_RT" | "Figure" | "Apptronik_Apollo" | "1X_RFM" | "Tesla_Optimus" | "BD_Atlas" | "Spot" | "ANYmal" | "ALOHA" | "WidowX" | "Franka" | "iiwa" | "Waymo" | "Wayve" | "Tesla_FSD" | "CARLA" | "MuJoCo" | "sim-only:<name>" | "other:<name>",
    "real_hardware_validated": true | false,
    "cross_embodiment_count": null | integer,
    "task_horizon_steps": null | integer,
    "novel_scene_pct": null | number,
    "vendor_self_eval": true | false,
    "third_party_reproduction": null | "<paper_id>" | "<lab_name>",
    "claimed_advantage_over_baseline": null | "<short claim>",
    "rebuttal_papers": [{"paper_id": "...", "summary": "..."}, ...],
    "notes": "..."
  },
  ...
]
```

## Bill classifier rule format

`scripts/bill_classifier.py` rules are tuples of (regex, bill, confidence-modifier):

```python
RULES = [
    # (pattern, bill, weight)
    (r"\bsim[\-_]?to[\-_]?real\b|\bSim2Real\b", "Bill_2", 1.0),
    (r"\bcross[\-_]?embodiment\b", "Bill_8", 1.0),
    (r"\b(VLA|vision[\-_]?language[\-_]?action)\b", "Bill_1_or_8", 0.85),
    (r"\b(tele[\-_]?op|tele[\-_]?operation|human[\-_]?in[\-_]?the[\-_]?loop)\b", "Bill_13", 0.95),
    (r"\b(long[\-_]?horizon|multi[\-_]?step\s+(plan|task))\b", "Bill_6", 0.95),
    (r"\b(MPC|model\s+predictive\s+control|whole[\-_]?body[\-_]?control|inverse\s+kinematics)\b.*\b(baseline|comparison)\b", "Bill_7", 0.90),
    (r"\b(METR|Apollo\s+Research|DARPA|NIST\s+robotics)\b", "Bill_10", 0.95),
    # ...
]
```

Two-pattern AND-form rules used for sub-bill resolution:

```python
RULES_AND = [
    # ((p1, p2), bill, weight)
    ((r"\bVLA\b", r"\b(real[\-_]?world|deployment|customer)\b"), "Bill_5", 0.9),
    ((r"\bsim2real\b", r"\b(camera|lighting|friction|gripper|contact)\s+(gap|model)\b"), "Bill_2", 1.0),
    ((r"\b(humanoid|biped)\b", r"\b(walking|running|locomotion)\b"), "Bill_11_locomotion", 0.95),
]
```

## Wiki paper file YAML frontmatter (`wiki/papers/<source>:<date>/<paper_id>.md`)

```yaml
---
paper_id: arxiv:2412.01234
title: "Helix: A Vision-Language-Action Model for Generalist Humanoid Control"
authors:
  - First Last
date: 2024-12
venue: arxiv:cs.RO 2024-12
summary: |
  Three-sentence summary tied to the closure mechanism.
bill: Bill_5
meta_cost: M3
verdict: known_bill
confidence: 0.85
watchlist_tier: quarterly
system_class: VLA
embodiment: Figure
real_hardware_validated: true
cross_embodiment_count: 1
task_horizon_steps: 12
novel_scene_pct: 15
vendor_self_eval: true
third_party_reproduction: null
claimed_advantage_over_baseline: "+22% on Figure-internal manipulation eval vs RT-2"
rebuttal_papers: []
schema_version: "1.0"
---

(body of the paper note: extended discussion, why this paper triggers Bill_5
specifically, what cousin papers are in the watch-list, etc.)
```

## Bill file format (`wiki/bills/bill_NN_<name>.md`)

```yaml
---
bill_number: 5
bill_name: "Sim-to-real generalization"
type: bill
empty_space_candidate: true
---

# Bill 5 — Sim-to-real generalization

## What gets paid

(prose explanation of the closure mechanism)

## Triggers

| Paper | Date | Notes |
| --- | --- | --- |
| [[arxiv:2412.01234]] | 2024-12 | ... |

## Cousin precedents

(historical anchor papers that established this closure mechanism)

## Re-poll cadence

monthly | quarterly | triggered

## Rebuttal literature

| Paper | Date | What it closes |
| --- | --- | --- |
| ... | | |
```

## Schema versioning

- `schema_version: 1.0` — current draft, will lock when batch 1 returns
- Future versions: bump on breaking change, document in `wiki/concepts/schema_changelog.md`