How to Audit a Catalyst Experimental Method Before Laboratory Work
A useful catalyst-method audit checks whether a preparation and evaluation plan is executable, reproducible, comparable, attributable, measurable, and ready for qualified safety review. This guide shows how the Audit Catalyst Experimental Methods Skill reviewed a synthetic Ni/Al₂O₃ CO₂ methanation plan before laboratory work began.
The example below comes from a completed Skill run. Its HTML, Word, and JSON outputs passed the same consistency check.
Ni/Al₂O₃ Methanation Catalyst Method Audit
The main method is identifiable, but the comparison and measurement plan cannot yet support a reliable attribution claim. The run found three material issues and no critical issue; this is not laboratory-safety approval.
What this supports
Revising the method before execution and creating an auditable list of required additions.
What this does not prove
That omitted work was not performed, that results are authentic, or that laboratory safety has been approved.
Supply the preparation steps, sample design, evaluation conditions, characterization plan, and claims to be supported.
What a catalyst method audit should check
A catalyst method audit asks whether a trained researcher can execute the stated operations, reproduce the conditions, compare samples fairly, attribute an observed effect to the intended variable, and connect each claim to suitable evidence.
The deliverable combines an overall judgment, priority issues, preparation-step register, sample-variable register, required additions, and claim-validation matrix. It is most useful before experiment time and material are committed.
Why method descriptions create false confidence
A familiar sequence — impregnation, drying, calcination, reduction, fixed-bed testing, and gas chromatography — can sound complete while omitting the values that control reproducibility. A technique name is not a complete measurement method, and a comparator is not automatically a valid control.
The audit must also preserve a crucial distinction: ‘not specified’ does not mean ‘not performed.’ The submitted document is being assessed for reviewability, not used to accuse the laboratory of skipping work.
The Skill packages a professional, repeatable audit method and keeps issue counts, required additions, and output files consistent. When follow-up research needs external patent or scientific context, Patsnap also provides structured research data with frequently updated coverage. This Sample audited supplied material only.
How the Skill audits the method
The Skill classifies the material, extracts canonical method elements, checks preparation executability, builds a sample-variable register, audits the measurement chain, and links performance, structural, and mechanistic claims to evidence classes.
Its local runner generates one JSON context, one HTML report, and one Word report. A separate validator reconciles issue counts and required sections across the outputs.
Prepare, install, and run
Provide the method in its current form: a plan, paper Methods section, patent example, screenshot transcription, or machine-generated route. Include preparation, evaluation, controls, characterization, replication, and safety information when available.
No MCP connection is required. The workflow is local and deterministic; external patent or scientific evidence is a separate optional research task and does not replace method-specific measurements.
Audit this catalyst preparation and evaluation plan for executability, reproducibility, fair comparison, attribution, measurement reliability, claim-to-evidence linkage, and safety-review readiness. Treat missing details as not specified rather than not performed, and return prioritized required additions.How to interpret and act on the audit
Resolve issues in order of decision impact. For this Sample, add a support-only control, define replication and uncertainty handling, and connect composition or structure claims to methods that can distinguish the proposed state from alternatives.
After revision, rerun the same audit and compare the issue register. “Not specified” identifies a documentation gap; it does not prove that work was not performed. Qualified laboratory and EHS reviewers still own operating limits, gas handling, containment, stop-work criteria, and approval.
Bring current evidence into your AI workflow
Connect your agent to patent, scientific, and engineering research tools when the next question requires verifiable external evidence.
Frequently asked questions
Does a clean audit prove the experiment is safe?
No. The audit prepares the method for qualified laboratory and EHS review; it does not grant approval.
Why does the report say ‘not specified’?
Because silence in a submitted method is not evidence that the laboratory did not perform the step.
Does the Skill require Patsnap data?
No. The audit runs locally on supplied material.
What files does it produce?
The deterministic runner produces a JSON context, an HTML report, and a Word report, followed by an output-validation step.
Disclosure: This Skill audits the completeness and reviewability of supplied catalyst methods. “Not specified” does not mean “not performed,” and the output does not authenticate experimental results, approve laboratory safety, or replace qualified scientific and EHS review.
