> For the complete documentation index, see [llms.txt](https://docs.hivel.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.hivel.ai/hybrid-hivel-agent/2.-getting-started.md).

# 2. Getting Started

Hivel-agent runs on **your** **infrastructure** and syncs engineering metadata from your tools to the Hivel platform. It connects to your Jira, GitHub, GitLab, Bitbucket, Azure DevOps and Claude, pulls metadata incrementally, and securely uploads it to Hivel for analysis.

**Only metadata is sent to Hivel - no source code, no file contents, no attachments, no analytics or telemetry, no raw credentials.**

Everything the agent sends to Hivel is under your control:

* **Metadata only** - issue fields, PR metadata, commit metadata, code quality measures. No code, no comments content unless explicitly included.
* **No hidden telemetry** - the agent sends nothing to Hivel beyond what is described in this document.
* **Agent logs are opt-in** - stream\_logs: true in config enables real-time log streaming to Hivel for support purposes. It is off by default. Logs never leave your network unless you explicitly enable this.
* **Config upload is opt-in** - send\_agent\_config: true uploads a sanitized copy of your config (all credentials redacted) to help Hivel support diagnose issues. Off by default.
* **Granular include/exclude** - filter by project, repository, branch, user, field, or file path. PII can be redacted or stripped before upload.

See Configuration Reference and PII and text redaction controls for the full options.

### What data is collected

All data collected is metadata about work, not the work product itself.

| **Source**   | **What is collected**                                                                                                                                            |
| ------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Jira         | Projects, issue types, statuses, priorities, resolutions, users, issues + changelogs, sprints, boards, worklogs (opt-in), custom field values you explicitly map |
| GitHub       | Organizations, repos, branches, commits, pull requests (with reviews and review comments), teams, members                                                        |
| GitLab       | Groups, projects, branches, commits, merge requests (with reviews and comments), members                                                                         |
| Bitbucket    | Workspaces, repos, branches, commits, pull requests, PR comments, members                                                                                        |
| Azure DevOps | Organizations, projects, repos, branches, commits, pull requests (with reviews and comments), teams                                                              |
| Claude       |                                                                                                                                                                  |

If you also run the optional work-metrics job (section-7), it additionally produces: a work-type classification per commit (new work / rework / maintenance / assistance) and added/removed line counts per merged pull request. These are derived numbers, not source content - see (section-7) for exactly how they're computed and what does or doesn't leave your network in the process.

**Text content**

Commit messages, PR/MR titles and descriptions, and review comment text are included by default, because they're needed for engineering workflow analytics. If you don't want free text leaving your environment at all, two config-level controls remove it - see (section-7).

**Jira field scope**

Jira sync always includes a fixed, standard field set: issue key, summary (title), status, issue type, priority, assignee/reporter/creator, dates, resolution, labels, parent/subtask/issue-link relationships, and project metadata. Issue descriptions and comment bodies are not part of this standard set. Only fields you explicitly map via extra\_fields in your config are added beyond it - so if you never map a "Description" or comment field, that text is never read or sent.

### What is explicitly never collected

These are not configuration defaults that happen to be off - there is no code path in the agent that reads or transmits any of the following, for any provider:

* **Source code or file contents.** The agent never reads a repository's file contents, never checks out a working tree of source for upload, and never transmits file bodies.
* **Diff or patch content.** For every provider (GitHub, GitLab, Bitbucket, Azure DevOps, Claude), the per-file change record captured for a commit or pull request contains only the file path, change status (added/modified/removed/renamed), and added/removed line counts. There is no field, anywhere in the agent, that carries the actual diff hunk or patch text.
* **Attachments.** Issue/PR/MR attachments are never downloaded.
* **Secrets, tokens, or credentials belonging to your systems.** These are read from your local .env file (or fetched via the Hivel-managed credential option, [section-7](/hybrid-hivel-agent/8.-work-metrics-credentials-and-controls.md)) to authenticate to your tools - they are never included in any upload payload.
* **Anything from a tool you haven't configured.** The agent has no discovery or default-scan behavior; it only ever contacts the specific Jira/GitHub/GitLab/Bitbucket/Azure DevOps/Claude instances named in your config.yml.

### [Network calls - inbound and outbound](/hybrid-hivel-agent/2.-getting-started/network-calls-inbound-and-outbound.md)

### [System requirements](/hybrid-hivel-agent/2.-getting-started/system-requirements.md)

<br>

{% columns %}
{% column %} <a href="/hybrid-hivel-agent.md" class="button primary" data-icon="backward">Hybrid (Hivel agent)</a>
{% endcolumn %}

{% column %}

{% endcolumn %}

{% column %} <a href="/hybrid-hivel-agent/3.-prerequisites-and-set-up.md" class="button primary" data-icon="forward">Prerequisites and Set-up</a>
{% endcolumn %}
{% endcolumns %}