Rank #02 · Weekly · Security
Sherlock: Hunt Usernames Across 400+ Social Networks
One command — sherlock user123 — probes hundreds of sites for that username and writes the hits to a text (or CSV/XLSX) file.
Canonical digest: /daily/sherlock (also on Weekly board).
What it does
Sherlock is a mature username OSINT tool: given one or more usernames, it queries a large site list (README: 400+ social networks) and reports where accounts appear to exist. GitHub description: "Hunt down social media accounts by username across social networks." Homepage: sherlockproject.xyz.
From docs/README.md and pyproject.toml (package sherlock-project):
- CLI entry
sherlockvia Poetry packaging; install withpipx install sherlock-project, Docker (sherlock/sherlock), or distro packages (Fedoradnf; community packages for Debian/Ubuntu/Homebrew/Kali/BlackArch — third-party packages are not maintained by the project). - Search one user (
sherlock user123) or many; results land in per-username text files (optional--csv/--xlsx/--folderoutput). - Flags cover site filters (
--site), proxy, timeout, JSON site data override, Tor-oriented socks via PySocks/stem deps, NSFW site opt-in (--nsfw), and browse-open results. - Similar-username expansion with
{?}placeholders (replace with_,-,.).
Responsible use. Sherlock is an open-source reconnaissance helper for legitimate OSINT: authorized investigations, security research, and checking your own digital footprint. Do not use it to stalk, harass, or access accounts you do not own. Respect site terms and local law. This write-up does not describe bypassing platform protections or unauthorized access.
Who it's for
- Security analysts, journalists, and investigators who need a first-pass username sweep across many public profiles.
- Individuals auditing how widely a handle they control (or plan to use) already appears.
- Educators teaching OSINT basics with a well-documented MIT CLI.
Less useful if you need deep profile scraping, private data, or authenticated API access — Sherlock checks existence-style signals against its site list, not full profile dumps.
Stars / language / license
112,490 stars · Python · MIT
Data verified: 2026-11-16 09:10 Asia/Shanghai (official Weekly Top10 gate; language/license confirmed via GitHub API get_repository + prior README / pyproject notes)
Trending note: +10,412 stars this week (weekly Trending; selection captured 09:10 Asia/Shanghai)
Category tags
Security · OSINT · username-search · CLI · Python · reconnaissance · infosec
How it differs from similar projects
Compared with one-off curl / browser checks. Sherlock batches hundreds of sites with a maintained data list and export formats.
Compared with commercial people-search / OSINT SaaS. Those often add phone/email/breach aggregation behind a subscription. Sherlock is self-hosted, username-focused, MIT, and free to run locally or in Docker.
Compared with Maigret / WhatsMyName-class tools. Same problem space (username → site hits). Pick based on site coverage freshness, packaging, and output format — TidyHow has not ranked accuracy or false-positive rates across tools.
Caveats. Site lists go stale; some community packages are broken (README warns ParrotOS / Ubuntu 24.04 third-party packages). False positives/negatives are expected when sites change HTML or block probes. No independent benchmark of hit rate is claimed here. This week's board placement is reported as Trending growth only — Sherlock holds the week's second-highest climb as a classic username OSINT tool adding five figures; we do not invent a causal news hook.