agentleFS
NVIDIAKnown publisherNVIDIA/Quantum-Calibration-Agent-Blueprint/data/knowledge/skills/experiment-execution/SKILL.md

experiment-execution

Skill69 starsChanged 7 months agoApache-2.0View on GitHub

Reference guide for running individual quantum calibration experiments and interpreting their results. Use when the user asks to run a single experiment (e.g. resonator spectroscopy, qubit spectroscopy, T1, T2), inspect experiment plots with the VLM, or query experiment history and schemas via the lab tool.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. TenureAI/PhD-Zero★ 52same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository5 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 10 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Display results: - ALWAYS include this tag in your response: <experiment experiment-id="THE_EXPERIMENT_ID" /> (Replace THE_EXPERIMENT_ID …

    line 65
  2. Distinctive:

    Verify success criteria: if defined

    line 70
  3. Distinctive:

    After running an experiment, verify results against success criteria when applicable.

    line 75
  4. Distinctive:

    DO verify: when success criteria are defined in the workflow plan

    line 78
  5. Distinctive:

    DO verify: when numeric thresholds need checking (SNR, linewidth, depth, etc.)

    line 79
  6. Distinctive:

    SKIP verification: if user has already seen the plot and confirmed success

    line 80
  7. Distinctive:

    SKIP verification: if no specific success criteria were provided

    line 81
  8. Very specific:

    Visual (VLM): Use vlm_inspect to analyze plots

    line 85
  9. Distinctive:

    Ask specific questions matching the criteria

    line 86
  10. Distinctive:

    Programmatic: Check returned data values directly

    line 90
  11. Distinctive:

    No visible peak/dip

    line 113
  12. Distinctive:

    If experiment fails, try these adjustments:

    line 121

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Experiment Execution Guide

Reference guide for running and analyzing quantum calibration experiments.

Tools Available

run_experiment

Execute a quantum calibration experiment.

run_experiment(
    experiment_name="resonator_spectroscopy",
    params={
        "center_freq": 6.0,
        "span": 0.2,
        "num_points": 101,
        "power": -20,
        "num_averages": 2000
    }
)

Returns: {"id": "...", "status": "success", "results": {...}, "plots": [...]}

vlm_inspect

AI-powered visual analysis of experiment plots.

vlm_inspect(
    experiment_id="20240315_103015_resonator",
    prompt="Is there a clear dip? What is the resonator frequency and estimated SNR?"
)

Returns: Text analysis of the plot.

lab

Query experiment information and history.

lab(action="schema", experiment_name="resonator_spectroscopy")  # Get parameter schema
lab(action="history_show", experiment_id="...")                  # Get experiment details
lab(action="get_stats", experiment_id="...", array_name="...")  # Get array statistics

⚠️ Data Analysis Rules

NEVER copy raw experiment arrays into shell commands. Large arrays will be truncated and cause syntax errors.

DO use these approaches:

  1. VLM analysis: vlm_inspect(experiment_id=..., prompt="...") - visual plot analysis
  2. Returned scalars: run_experiment returns fitted values like resonator_freq, t2_star, etc.
  3. Lab stats: lab(action="get_stats", experiment_id=..., array_name="...") for array statistics

DON'T do this:

# BAD - will fail with syntax errors
execute("python3 -c 'data = [0.1, 0.2, 0.3, ...]'")  # Arrays too long!

Standard Execution Flow

  1. Run experiment with appropriate parameters
  2. Get experiment_id from result
  3. Display results - ALWAYS include this tag in your response: <experiment experiment-id="THE_EXPERIMENT_ID" /> (Replace THE_EXPERIMENT_ID with the actual ID. This shows the interactive viewer with plots.)
  4. Analyze plots with vlm_inspect - ask about peaks, dips, resonances, SNR
  5. Extract values from VLM analysis (frequency, linewidth, SNR)
  6. Verify success criteria if defined
  7. If failed, adjust and retry (up to max attempts)

Verifying Success Criteria

After running an experiment, verify results against success criteria when applicable.

When to Verify

  • DO verify when success criteria are defined in the workflow plan
  • DO verify when numeric thresholds need checking (SNR, linewidth, depth, etc.)
  • SKIP verification if user has already seen the plot and confirmed success
  • SKIP verification if no specific success criteria were provided

Verification Methods

Visual (VLM): Use vlm_inspect to analyze plots

  • Ask specific questions matching the criteria
  • Example: "Is there a single clear peak? What is the SNR? Is it greater than 5?"
  • Example: "Is there a visible dip with depth > 3 dB?"

Programmatic: Check returned data values directly

  • Extract values from experiment results (e.g., snr, peak_frequency, linewidth)
  • Compare against thresholds in success criteria

Typical Parameter Ranges

ExperimentParameterTypical Range
qubit_spectroscopystart/stop4500-5500 MHz
qubit_spectroscopystep0.5-10 MHz
qubit_spectroscopynum_avs500-5000
resonator_spectroscopystart/stop6000-8000 MHz
resonator_spectroscopystep0.1-1 MHz

Success/Failure Indicators

Good Data (Success)

  • Clear single peak or dip visible
  • SNR >= 5 (ideally >= 8)
  • Peak width reasonable (5-50 MHz for qubit)
  • Smooth baseline, low noise

Bad Data (Retry Needed)

  • No visible peak/dip
  • Multiple peaks (ambiguous)
  • SNR < 3
  • High noise floor
  • Saturation or clipping

Retry Strategies

If experiment fails, try these adjustments:

  1. No peak found: Expand frequency range (1.5x wider)
  2. Low SNR: Increase num_avs (2x more averaging)
  3. Multiple peaks: Narrow range around strongest peak
  4. High noise: Increase num_avs, check for interference

Example Execution

Task: Run resonator spectroscopy centered at 6.0 GHz

Attempt 1:
> run_experiment(experiment_name="resonator_spectroscopy",
                 params={"center_freq": 6.0, "span": 0.2, "num_points": 101, "num_averages": 2000})
> Result: experiment_id = "20240315_103015_resonator"
> Include in response: <experiment experiment-id="20240315_103015_resonator" />
> vlm_inspect(experiment_id="20240315_103015_resonator",
              prompt="Is there a clear dip? What is the frequency and SNR?")
> VLM: "Clear dip visible at 5.823 GHz. Estimated SNR ~12. Clean baseline."

Success! Extracted: resonator_frequency = 5.823 GHz

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.