# Aptlantis Analysis Standard

Structures local analysis runs, metrics, evidence, and interpretation limits.

Catalog: active · v0.2.1 · reviewed 2026-09-23

Source maturity: candidate · Explanation reviewed 2026-10-01

Applies to: Analysis and evaluation runs that need reproducible inputs, tools, assumptions, metrics and evidence.

[Public page](https://aptlantis.net/city-hall/aas)

<a id="purpose-and-applicability"></a>

## Purpose and applicability

An unexplained score cannot support a decision. AAS records what was evaluated, how a metric was computed and how far the finding can be interpreted.

Analysis and evaluation runs that need reproducible inputs, tools, assumptions, metrics and evidence.

<a id="how-it-works"></a>

## How it works

Create an analysis manifest for setup, then a run record for a specific execution. Name input versions, environment, commands/procedure, metrics with units/denominators/exclusions, outputs and baseline. Separate facts, inferences and assumptions.

Outputs are the manifest, run evidence, preserved metrics, interpretation, limitations and follow-up questions. Carry DDS provenance and split uncertainty into the findings.

<a id="in-practice"></a>

## In practice

This scaffold evaluates required suite-artifact coverage. It defines a metric but records no score. Even a real pass would support a presence finding, not a claim that the standards are semantically complete.

### Evaluation-run scaffold

illustrative · teaching example, not a verification result

Adapted run record with assumptions and a deliberately absent finding.

Source: `AAS/examples/Example-Evaluation-Run-Record.md`

```text
Intent: inspect required suite-artifact coverage.
Input: approved synthetic suite inventory, version to be recorded.
Method: compare declared references with present artifacts.
Metric: present required references / total required references.
Unit: fraction; denominator: required references in the declared scope.
Exclusions: domain compliance and prose quality.
Assumption: the required-reference list is correct.
Tools/environment: pending.
Observed findings: none; run not performed.
Interpretation limit: presence alone cannot establish maturity.
Follow-up: execute against an approved snapshot and preserve output.
```

[Inspect Evaluation-run scaffold](/city-hall/standards/aas/example-1.txt)

<a id="adopt-one-part"></a>

## Adopt one part

Start with a bounded surface or record. Complete the relevant adopter checks before extending the claim.

1. Define one evaluation question and name the exact input snapshot and comparison baseline.

2. Define the metric, denominator, exclusions, tools and procedure before interpreting the result.

3. Preserve actual output and distinguish observed findings from assumptions; carry source limitations forward.

<a id="sources-and-limits"></a>

## Sources and limits

No evaluation was run for this specimen. Presence checks do not measure prose quality or standard maturity. Placeholder outputs and missing provenance must remain limitations, not be converted into findings.

These are reviewed public explanations, not the normative specifications. Suite references are relative to the canonical collection; site/ references identify committed website sources and webserver/ references identify serving configuration. Illustrative examples demonstrate record shape; they do not establish compliance. Suite checks and adopter validation are separate.

- `AAS/AAS.manifest.toml`
- `AAS/Adoption-Guide.md`
- `AAS/Validation-Checklist.md`
- `AAS/Aptlantis Analysis Standard.md`
- `AAS/examples/Example-Evaluation-Run-Record.md`
- `AAS/AnalysisManifest.schema.toml`

[Reviewed manifest facts](/city-hall/components/aas.md)

## Related responsibilities

- [Dataset Development Standard](/city-hall/dds): Owns provenance, rights and splits of input datasets.
- [Agent Task Standard](/city-hall/ats): Records task actions and handoff separately from analytical evidence.
- [Standards Framework Development Standard](/city-hall/sfds): Defines suite structure, not the analysis result.