DocSlicer collects
nothing.

This policy applies to the DocSlicer software — the Python library, the command-line tool, and the MCP server — and to this website at docslicer.ai. It is published by the maintainers of DocSlicer, who act as the data controller for the small amount of information described in sections 07 and 08. Everywhere else in this document, “DocSlicer” means the software running on your own hardware.

The source code is public at github.com/DocSlicer/DocSlicer. Everything claimed here can be verified by reading it.

Nothing. DocSlicer has no telemetry, no analytics, no crash or error reporting, and no usage tracking. It does not require an account, a licence key, an API key, or registration, and it has no mechanism for sending us information about you, your machine, or your documents — not on install, not on first run, not on update, and not on error.

This section is about the software. Visiting this website is a separate matter, described in section 07: the site does run analytics. Nothing collected there is connected to the software, which has no way to report anything in the first place.

We run no advertising, build no profiles for it, make no automated decisions about you, and sell or share personal information with no one.

Parsing runs entirely on your own machine, in a local process. Document contents are used only to produce the outline, text slices, search results, and markdown you ask for, and are returned only to the caller that asked for them. Documents are never uploaded to DocSlicer or to any third party, and are never used to train any model.

When running as an MCP server, DOCSLICER_MCP_ROOT bounds which directory tree may be read from and written to. Paths outside that tree are rejected.

Parsed results are cached on your own disk — by default ~/.cache/docslicer-mcp, configurable with DOCSLICER_MCP_CACHE. The cache is pruned to a size ceiling (DOCSLICER_MCP_CACHE_MAX_MB, default 2048 MB); otherwise it persists until you delete it.

Deleting that directory removes the cached data permanently, with no copy retained anywhere else — there is no backup, no sync, and no remote store. Nothing is written outside the cache directory and any output path you supply.

No network request is made for a local file. Requests leave your machine only when you pass an http(s) source: that URL is fetched directly from the host you named, and for HTML pages Playwright may load the subresources that page references, exactly as a browser would.

Requests to sec.govsend a User-Agent header identifying the client, as the SEC fair-access policy requires. These hosts are third parties chosen by you, not by DocSlicer; they will see your IP address and the request as any web client's would appear, and their own privacy policies govern what they log. Set DOCSLICER_MCP_ALLOW_URLS=0 to reject remote sources entirely.

When DocSlicer runs as an MCP server, the client you connect it to — Claude, Cursor, or any other MCP-capable tool — handles the conversation under its own privacy policy, including whatever document text it chooses to send to its own model provider. DocSlicer is not a party to that arrangement and receives nothing from it. If the confidentiality of a document matters, check the policy of the client you are using.

docslicer.ai is a marketing and documentation site, and it runs analytics. We use two services: Google Analytics and PostHog. We use them for an unglamorous reason — to learn which documentation pages people actually read, and which ones lose them — and there is no sign-up on this site, so nothing collected is tied to a name or an email address.

Between them they record: the pages you view and the order you view them in, the site or search engine that referred you, your approximate location derived from your IP address, your browser, operating system, and screen size, and which links and buttons you click. PostHog also records a playback of your visit — the pages, clicks, scrolling, and navigation, reconstructed from those events rather than filmed. Text you type into form fields is masked before it leaves your browser.

Both set cookies to recognise a returning browser across pages and visits: Google Analytics sets _ga and _ga_*, and PostHog sets cookies beginning ph_. These are analytics cookies, not advertising ones. We run no ad network on this site, serve no tracking pixels for one, and do not use this data to follow you to other websites.

This data is processed on our behalf by Google and by PostHog, on servers in the United States, and is retained for the period configured in each service (Google Analytics defaults to 14 months). Their own privacy policies, linked above, govern what they do with it as processors. We do not sell it, and we do not combine it with anything else.

Refusing it. Any content blocker or privacy-focused browser will stop both scripts loading, and the site is built to work perfectly well without them — nothing here depends on analytics running. Google also publishes a browser opt-out add-on covering every site that uses Google Analytics. If you would rather we deleted what has already been recorded for your visits, email us and we will.

Separately from analytics, the site is served by a hosting provider that, like any web server, processes standard request data — IP address, user agent, requested URL, timestamp — to deliver pages and protect against abuse. Those logs are short-lived operational records held by the host under its own policy; we do not query them to identify visitors. Pages that show the GitHub star count fetch that number server-side, so your browser makes no request to GitHub for it.

Links from this site to GitHub, PyPI, and other third parties are governed by those services' own privacy policies once you follow them.

If you email us or open a GitHub issue, we necessarily receive what you send — your address or username, and whatever you write. We use it only to answer you, and we keep it no longer than that correspondence needs. Please do not send confidential documents by email; a redacted reproduction is almost always enough to debug a parsing problem.

GitHub issues are public by design. Anything you post there is visible to everyone and governed by GitHub's privacy policy as well as ours.

Privacy laws including the GDPR and the CCPA give you rights to access, correct, export, and delete personal data held about you, and to object to its processing. For users of the software those rights are close to moot: we hold no personal data about you, so there is nothing to export or erase. Data produced by your use of DocSlicer lives on your own disk and is under your control alone — deleting the cache directory is a complete erasure.

Where we do hold something, the rights are real and we will honour them. That means the website analytics in section 07 and the correspondence in section 08: ask us at any time what we hold and to delete it, and we will. Section 07 also explains how to stop the analytics collecting anything further. We do not sell or share personal information, and never have. If you are in the EU or UK and believe we have handled something badly, you may also complain to your local data protection authority — though we would rather you told us first.

DocSlicer is a developer tool, not directed at children, and it knowingly collects no information from anyone, including children under 13 (or 16 where local law sets that bar).

The strongest security property here is architectural: your documents never travel to us, so there is no server of ours that can leak them. The security of the parsed cache and any output files is a function of your own machine and filesystem permissions.

If you find a vulnerability in DocSlicer, please report it privately by email rather than in a public issue, and we will work with you on a fix and a disclosure timeline.

If this policy changes, the updated version is published on this page with a new “last updated” date, and material changes are noted in the project changelog. Because we collect no contact details for users of the software, this page is the only place we can announce a change — the previous wording remains visible in this site's git history.

Privacy questions: hello@docslicer.ai. Bugs and feature requests: GitHub Issues.