> ## Documentation Index
> Fetch the complete documentation index at: https://docs.corridor.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Export Team Data

> Download a ZIP of your team's projects, members, guardrails, findings, PR reviews, MCP servers, and compliance data.

A team Owner or an organization owner can download a copy of the team's data as a ZIP of CSV files. The export holds what your team can see in the Corridor dashboard.

## Who can export

Only a team **Owner** or an organization owner can start an export and download it. An organization owner sees the card even when their team role is Admin. Admins, Standard Users, Viewers, and Security Reviewers do not see the card. The server applies the same rule. API tokens cannot start or download an export.

## Start an export

1. In the sidebar, click **Team Settings**.
2. Scroll past the team card to the **Export team data** card, near the bottom of the page.
3. Click **Start export**.

The export runs in the background. A small team takes seconds, a large team can take a few minutes. You can leave the page. The table on the card lists your team's last 10 exports. A running export shows its current step, for example "Exporting findings (4 of 7)".

When the status is **Completed**, click **Download**.

## Limits

* One export runs at a time per team. While one runs, **Start export** is disabled.
* A team can start at most 5 exports in 24 hours.
* An export that runs for 12 hours stops and shows **Failed**.
* If an export fails, start a new one. If it fails again, contact support with the export ID shown on the card.

## Download link and retention

* Each click on **Download** creates a new download link. The link works for one hour.
* Corridor deletes the ZIP 7 days after the export completes. The status then shows **Expired**, and the ZIP cannot be downloaded. Start a new export to get current data.

## What the ZIP contains

The ZIP holds the CSV files below and a `manifest.json` file. Each CSV file starts with a header row. Dates are ISO 8601 in UTC.

| File | Data | Time range |
| - | - | - |
| `projects.csv` | Your team's projects | Current state |
| `members.csv` | Team members and their roles | Current state |
| `guardrails.csv` | Guardrails of every project, archived ones too | Current state |
| `findings.csv` | Findings shown on the Findings page | All time |
| `pr_reviews.csv` | PR reviews shown on the PR Reviews page | All time |
| `pr_review_comments.csv` | Comments of those PR reviews | All time |
| `pr_review_findings.csv` | Findings raised in those PR reviews | All time |
| `mcp_servers.csv` | MCP servers detected for your team's members | Last 60 days |
| `user_compliance.csv` | Each member's IDE extension compliance status | Current state |
| `manifest.json` | A summary of the export | The moment of the export |

There is no option to choose a time range. "All time" means every row the dashboard shows, at any age.

### The export is not a snapshot

Corridor reads each file in turn while the export runs. Data that changes during the export can be in one file and not in another.

* `findings.csv`, `pr_reviews.csv`, and `mcp_servers.csv` hold only rows created up to the moment you clicked **Start export**.
* `projects.csv`, `members.csv`, `guardrails.csv`, and `user_compliance.csv` hold what is stored when the export reaches that file. They can include rows created after you clicked.
* `pr_review_comments.csv` and `pr_review_findings.csv` hold every comment and finding of the reviews in `pr_reviews.csv`. A comment or finding added to an older review after you clicked can be included.

### Feature flags

Some data shows in the dashboard only when a feature is on for your team. The export follows the same settings:

* If AI authorship is off for your team, `pr_reviews.csv` has no `ai_generated_percentage` column.
* If duplicate detection is on for your team, `findings.csv` and `pr_review_findings.csv` leave out findings marked as duplicates, as the dashboard does.
* If your team does not see Bugcrowd findings in the dashboard, `findings.csv` has none.

## Files and columns

### `projects.csv`

Every project of your team, except projects that are being deleted.

| Column | Description |
| - | - |
| `id` | Project ID |
| `name` | Project name |
| `repo_url` | Repository URL |
| `default_branch` | Default branch |
| `status` | Project status |
| `created_at` | When the project was created |

### `members.csv`

The members shown on the Teams page. Service accounts are not listed.

| Column | Description |
| - | - |
| `user_id` | User ID |
| `email` | Email address |
| `name` | Display name |
| `role` | The role the Teams page shows: Owner, Admin, Standard User, Viewer, or Security Reviewer |

### `guardrails.csv`

The guardrails of every project, as the project guardrails page shows them. Archived guardrails are included. Guardrails of projects that are being deleted are not.

Each row names its project, as the team guardrails page does.

| Column | Description |
| - | - |
| `id` | Guardrail ID |
| `project_id` | ID of the guardrail's project |
| `project_name` | Name of the guardrail's project |
| `name` | Guardrail name |
| `cwe` | CWE of the guardrail's template, if it has one |
| `guardrail_text` | The guardrail text |
| `type` | Guardrail type |
| `created_at` | When the guardrail was created |
| `updated_at` | When the guardrail last changed |
| `archived_at` | When the guardrail was archived; empty if active |

### `findings.csv`

The findings shown on the Findings page, at any age. Findings that Corridor's AI closed as false positives are not included, as on the Findings page. A finding closed by Corridor's false positive review is included. While a project is being deleted, its findings are left out, as in `projects.csv`.

| Column | Description |
| - | - |
| `id` | Finding ID |
| `title` | Title |
| `description` | Description |
| `enhanced_description` | Extended description, if Corridor added one |
| `affected_file` | File the finding is in |
| `vulnerable_excerpt` | The code excerpt |
| `severity` | Severity |
| `cwe` | CWE |
| `state` | State, for example open or closed |
| `closed_reason_category` | Why the finding was closed, as a category |
| `closed_reason` | Why the finding was closed, as text |
| `source` | Where the finding came from: `corridor` or `bugcrowd` |
| `source_identifier` | The finding's ID in that source |
| `closed_by_ai` | `true` if Corridor's AI closed the finding, including a detected fix; `false` if it did not, which covers open findings and findings a person closed; empty if no value is stored |
| `reachable_path` | `true` if Corridor found a reachable path to the code |
| `remediation_detected_at` | When Corridor detected a fix |
| `remediation_detected_in_commit` | Commit that fixed the finding |
| `remediation_confidence` | Corridor's confidence in the fix |
| `remediation_reasoning` | Why Corridor thinks the finding is fixed |
| `remediation_last_checked_commit` | Last commit Corridor checked for a fix |
| `project_id` | Project ID |
| `project_name` | Project name |
| `team_id` | Team ID |
| `team_name` | Team name |
| `rule_id` | ID of the guardrail that raised the finding, if any |
| `rule_name` | Name of that guardrail |
| `created_at` | When the finding was created |
| `updated_at` | When the finding last changed |

### `pr_reviews.csv`

The PR reviews shown on the PR Reviews page, at any age. Reviews that failed to post are not included.

| Column | Description |
| - | - |
| `id` | PR review ID |
| `task_id` | ID of the review run |
| `project_id` | Project ID |
| `project_name` | Project name |
| `team_id` | Team ID |
| `team_name` | Team name |
| `github_pr_id` | Pull request number, or merge request IID on GitLab |
| `github_pr_author_username` | Pull request author |
| `github_pr_review_url` | Link to the review on the pull request |
| `head_sha` | Commit that was reviewed |
| `event` | The review result: `COMMENT`, `REQUEST_CHANGES`, or `APPROVE` |
| `shadow` | `true` if the review ran in shadow mode and was not posted |
| `ai_generated_percentage` | Share of the changed code written by AI. Only when AI authorship is on for your team |
| `created_at` | When the review was created |
| `updated_at` | When the review last changed |
| `body` | The review text |

### `pr_review_comments.csv`

The comments of the PR reviews in `pr_reviews.csv`.

| Column | Description |
| - | - |
| `pr_review_id` | PR review ID |
| `comment_index` | Position of the comment in the review |
| `comment` | The comment, as JSON |

### `pr_review_findings.csv`

The findings raised in the PR reviews in `pr_reviews.csv`, as the PR review page shows them. The file uses the same state and duplicate rules as `findings.csv`. Four things differ:

* Findings that Corridor's AI closed as false positives are kept, as on the PR review page. For those rows, `false_positive_reason` says why.
* Findings of every source on those reviews are kept, including Bugcrowd.
* There is no date cutoff. A finding added after you clicked **Start export** is here when its review is older.
* Findings of a project that is being deleted are kept. `findings.csv` leaves them out while the deletion runs.

| Column | Description |
| - | - |
| `pr_review_id` | PR review ID |
| `finding_id` | Finding ID |
| `finding_title` | Title |
| `finding_cwe` | CWE |
| `finding_severity` | Severity |
| `finding_state` | State |
| `finding_affected_file` | File the finding is in |
| `false_positive_reason` | Why the finding was closed as a false positive |
| `remediation_detected_at` | When Corridor detected a fix |
| `remediation_detected_in_commit` | Commit that fixed the finding |
| `remediation_confidence` | Corridor's confidence in the fix |
| `remediation_reasoning` | Why Corridor thinks the finding is fixed |
| `remediation_last_checked_commit` | Last commit Corridor checked for a fix |

### `mcp_servers.csv`

The MCP servers detected for your team's members in the last 60 days. The 60 days end when you click **Start export**. This matches the **MCP Servers Detected** card on the Posture dashboard set to **Last 60 days**. The card shows the last 30 days by default. There is one row per server name, detection source, and status. A member of two teams has their servers counted in both teams' exports.

| Column | Description |
| - | - |
| `server_name` | MCP server name |
| `users` | Number of members who use the server |
| `detection_source` | How Corridor detected the server |
| `status` | The server's status |
| `last_seen_at` | When Corridor last saw the server |

### `user_compliance.csv`

The compliance status of each member, one row per member and platform. These are the rows the Governance **Members** page shows.

| Column | Description |
| - | - |
| `id` | Record ID |
| `user_id` | User ID |
| `user_email` | Email address |
| `team_id` | Team ID |
| `team_name` | Team name |
| `platform` | IDE or agent platform, for example `cursor` |
| `corridor_version` | Version of the Corridor extension |
| `value` | JSON with the details the Members page shows. For `cursor`: `licenseType`, `privacyModeEnabled`, `dbReadError`, and `lastCheckAt`. For other platforms: `lastCheckAt` |
| `created_at` | When the record was created |
| `updated_at` | The newest update to this member's records for the platform |

### `manifest.json`

One JSON object that describes the export.

| Key | Description |
| - | - |
| `profile` | Always `customer` |
| `version` | Format version, currently `1` |
| `team` | The team's `id` and `name` |
| `requestedBy` | Email of the person who started the export |
| `requestedAt` | When the export was started. `findings.csv`, `pr_reviews.csv`, and `mcp_servers.csv` hold only rows created up to it |
| `generatedAt` | When the ZIP was built |
| `files` | Each file's `name`, its number of `rows`, and its size in `bytes` |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.