Physics Plausibility Check
**What it does** Paste text, a robotics data record, or a network graph — get a deterministic PASS/FAIL with rule-by-rule evidence.
Free
Works with the AI Agents you already use
Physics Plausibility Check
Example session with this skill installed
Check this claim for physical plausibility: The experimental coolant reached a temperature of -5 Kelvin while the drone accelerated to a constant speed of 450,000 km/s.
- Read your context and instructions
- Compiled the physics plausibility check
I have audited the numeric claims regarding temperature and velocity using the physics plausibility engine. Both values violate fundamental physical laws.
Physics Audit Results
| Metric | Value | Verdict | Violation Detail |
|---|---|---|---|
| Temperature | -5 K | FAIL | Absolute zero breach: Temperature cannot be below 0 Kelvin. |
| Velocity | 450,000 km/s | FAIL | Special relativity violation: Exceeds the speed of light (c ≈ 299,792 km/s). |
| Mass/Energy | N/A | PASS | No mass-energy equivalence violations detected in text. |
| Probability | N/A | PASS | No statistical range errors detected. |
Next steps
- Correct the temperature scale: Ensure all cryogenic values are reported in Kelvin and remain strictly positive.
Adjust velocity parameters: Recalculate the drone's speed to ensure it stays within sub-relativistic limits for atmospheric flight.
3.
Verify units: Confirm if the intended temperature was 5 K or if a different unit (Celsius/Fahrenheit) was erroneously converted.
Connects securely to your tools. The creator never sees your data.
About this skill
What it does
Paste text, a robotics data record, or a network graph — get a deterministic PASS/FAIL with rule-by-rule evidence. No LLM inside, no opinion: pure physics rules. Same input always gives the same verdict (reproducibility std = 1.3e-15).
Six check families
- Numeric claims — speeds above light speed, temperatures below absolute zero, probabilities outside [0,1], negative masses or distances
- Kinematics — distance / speed / time must agree (v·t = d within tolerance)
- Conservation — before/after values in your text must balance (energy, mass, counts)
- Unit sanity — cm vs m vs km confusion, °C vs K, impossible conversions
- Robotics records — quaternion normalization, timestamp monotonicity, NaN/Inf scan, frame gaps, unit annotations — this rule set flagged real issues in public datasets from 4 robotics vendors
- Networks (paid API) — four HSEC predictions on graphs: phase transition, causal cone, curvature singularities, critical failure size
Real session
Buyer: "Check this claim: the drone flew at 500,000 km/s and returned in 2 minutes." Skill: FAIL — speed_exceeds_light (c = 299,792,458 m/s; 500,000 km/s is 1,668× c). Location and evidence included.
Free vs paid
- Free: rule checks 1–5 run locally on your machine. No network, no data leaves you.
- Paid (4.9 once): unlocks the server-side four-prediction engine (6) via a simple API call — the deep check used in our public audits of robotics datasets.
Honest boundary
A physics sanity gate — necessary, not sufficient. PASS means "no law-of-physics violation found by these rules", not "the claim is true". Every verdict states this.
Reproducibility
Every verdict carries rule IDs and locations so anyone can recompute it. Deterministic engine.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Necessary but not sufficient for correctness. Rule-based detection means unknown formats may pass silently. Excludes proprietary deep checks found in the full PHM engine.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 18 days ago
- Free with an Agensi account
- Software
Python 3.8+phm-lite-check
- Optional
- MCP host
Install with pip install git+https://github.com/PHM-Audit/phm-lite-check. The MCP host can be Claude Desktop or Cursor.