# 📜 Execution `metadata_provo.xml`: Execution Provenance Metadata

Among the metadata automatically generated during a CCP method execution, the `metadata_provo.xml` file plays a central role in ensuring **traceability**, **reproducibility**, and adherence to the **FAIR principles** (Findable, Accessible, Interoperable, Reusable).

This file is created **at runtime** and captures a structured record of the execution process using the [W3C PROV-O ontology](https://www.w3.org/TR/prov-o/), extended with CCP-specific terms (`d4s:` namespace).

## Purpose and FAIR Alignment

The `metadata_provo.xml` document provides a detailed and machine-readable description of:

- **What** was executed: method name, version, and parameters
- **Which infrastructure** and runtime (Docker image) was used
- **Who** triggered the execution (user and agent)
- **Where** the execution took place (VRE context)
- **When** it started and ended
- **What** was produced (standard output, error, results)

This aligns with the FAIR principles by:

- Ensuring **findability** of methods and outputs through persistent identifiers
- Promoting **accessibility** via downloadable and interoperable XML
- Supporting **interoperability** through PROV-O semantics
- Enabling **reuse** through complete and contextual provenance information

## File Availability

- Generated automatically with each execution
- Included in the `output.zip` if archiving is enabled
- Stored and referenced in archived executions in the CCP interface

## Structure Overview

The XML file contains:

- `<prov:activity>`: The execution session, with start/end time, operator name, version, status, and the triggering agent
- `<prov:entity>`: Each input, output, parameter, and resource is represented as an entity, with detailed types and descriptions
- `<prov:person>` and `<prov:softwareAgent>`: The user and system initiating the process

Each `<prov:entity>` includes:

- A `prov:value` indicating its actual content or reference
- A link to the `activity` that used or generated it
- Type annotations such as `d4s:IMPORTED`, `d4s:COMPUTED`, `d4s:text/plain`, etc.

## 🧪 Example Snippet

```xml
<prov:activity id="activity-...">
  <prov:startTime>2025-03-25T09:40:54.797Z</prov:startTime>
  <prov:endTime>2025-03-25T09:50:30.136Z</prov:endTime>
  <prov:type xsi:type="xsd:QName">d4s:computation</prov:type>
  <prov:softwareAgent prov:id="d4s:itineris.d4science.org"/>
  <prov:person prov:id="d4s:alfredo.oliviero"/>
  <prov:entity prov:id="d4s:operator_name">
    <prov:value>GDAL - Raster TIFF Details - Workspace</prov:value>
  </prov:entity>
</prov:activity>
```

## 💡 Note Notes

- The file is **not meant for manual editing**
- It is especially useful in **automated workflows**, **scientific validation**, and **long-term preservation**
- Fields such as `d4s:status`, `d4s:VRE`, and `d4s:operator_description` offer additional context

> ✅ CCP-generated provenance allows external tools, auditors, or collaborators to reconstruct not only **what** was done, but also **how**, **where**, and **by whom** — making it a key enabler of transparent and reproducible science.

## 🔗 Related Pages

- 📑 [Generated metadata files](./04_metadata_files)

- ⏯️ [Understanding executions](./01_executions)

- 📂 [Execution outputs and logs](./02_outputs_and_logs)

- 🌀 [Execution status and lifecycle](./03_execution_status)

- 🔗 [W3C PROV-O Specification](https://www.w3.org/TR/prov-o/)