Skip to content
KitploitKITPLOIT
ToolsExploitsBlog
Log in
Submit
ToolsExploitsBlog
Submit

Hacking, PenTest, and Cybersecurity Tools for Your Security Arsenal!

Kitploit is a directory of hacking, cybersecurity, and pentesting tools. Discover the latest project updates to find vulnerabilities, analyze systems, automate testing, and strengthen your security.

FeedsContactPrivacy© 2026 Kitploit

Tool Directory

Categories

View all categories
Loading categories
oss-scanner — Service that builds enrolled open-source repositories in isolated offline VMs and scans them for security vulnerabilities, emailing findings with reproducers and proposed patches. | Kitploit
Tools/GitHubGitHub/anthropics/oss-scanner
Defensive ToolsStatic AnalysisVulnerability ScannersDynamic Analysis (Sandboxing)Vulnerability AnalysisCode AnalysisSecurity VirtualizationDevSecOpsSupply Chain SecurityAI Security
GitHub
1.8k1.2k1507h 15m agoReviewed by Kitploit
Share
anthropics/oss-scanner

oss-scanner

Service that builds enrolled open-source repositories in isolated offline VMs and scans them for security vulnerabilities, emailing findings with reproducers and proposed patches.

View Repository

OSS Scanner

OSS Scanner is a service by Anthropic to scan critical open-source repositories for security vulnerabilities. Learn more about the service at red.anthropic.com/oss-scanner.

Note, use of this tool is subject to the terms and conditions outlined in the OSS Scanner terms.

How to sign up

Enroll an open-source project by opening a pull request that adds one directory:

projects/<name>

Our security scanner builds your project in an isolated VM, then analyses it with no Internet access and emails what it finds to primary_contact (and any CCs), each with a reproducer and a proposed patch where available. Reports are model-generated and are not reviewed by a human. Because of this we do not place a 90-day disclosure period on these findings and will not make them public. Project owners should provide project.yaml (to configure the scanner), a Dockerfile (stating how to build the project), and optionally a threat_model.md (providing project-specific threat modeling), a report_guidelines.md and a patch_guidelines.md (saying how you would like reports and patches to look).

project.yaml

Start from templates/project.yaml:

repo: https://github.com/example/project    # required: the git repository to scan; add #branch to pin one
primary_contact: [email protected]       # required: reports and build problems go here (one address)
auto_ccs:                                   # optional: more addresses on every mail
  - [email protected]
homepage: https://example.org               # optional
disabled: false                             # optional: true pauses reports without removing the enrolment
dockerfile: .oss-scanner/Dockerfile         # required unless a Dockerfile sits in this repo next to your project.yaml
threat_model: .oss-scanner/threat_model.md  # optional; also supports placing the file in this repository
# report_guidelines: .oss-scanner/report_guidelines.md  # optional; the same
# patch_guidelines: .oss-scanner/patch_guidelines.md    # optional; the same

repo and primary_contact are always required. dockerfile is required unless you keep your Dockerfile next to project.yaml (see below). Others are optional.

The email addresses in project.yaml are public. Use addresses you are happy to see published, such as a security alias.

To receive email reports encrypted, add your armored OpenPGP public key. They then go to primary_contact only; pgp cannot be combined with auto_ccs:

pgp: |
  -----BEGIN PGP PUBLIC KEY BLOCK-----
  mQINBF...
  -----END PGP PUBLIC KEY BLOCK-----

To choose where reports go, set the lowest severity each channel receives: low, medium, high, critical, or none to turn the channel off. A channel you leave out uses the default shown here, and a report at or above both minimums goes to both channels. pgp only applies to email, so GitHub reports are not encrypted to your key:

output_format:
  email:
    min_severity: low
  github_security_advisory:          # GitHub private vulnerability reporting
    min_severity: none

To turn on GitHub reports, first enable private vulnerability reporting on your repository and add oss-scanner as a trusted reporter. Anthropic operates the oss-scanner account, which our reports come from; without this, GitHub limits how many it can send you. We ask you to confirm it on your pull request before we merge.

You must provide a Dockerfile, in exactly one of two places:

  • in your own repository: set dockerfile: to its path (we suggest at some location like .oss-scanner/Dockerfile). This is the preferred option because it allows you to update the build without a pull request here; or,
  • here, next to project.yaml, as projects/<name>/Dockerfile, with no Dockerfile: key set in the project.yaml. If you would rather not add files to your repository, put it here and the scanner builds it exactly as it would an in-repo copy.

(The threat model works the same way. Set the threat_model field for its location in your repository, or place it here at projects/<name>/threat_model.md. So do the report guidelines and the patch guidelines, with the report_guidelines and patch_guidelines fields.)

The purpose of the Dockerfile is to set up your environment, install every dependency, and build the project. Initial project setup runs with network access enabled, but the security audit after it runs without Internet access. Anything the build or the tests need must be fetched during the initial Dockerfile setup. We recommend checking that your tests pass inside the built image.

The threat_model.md (optional but strongly recommended) file allows you to provide the scanner with documentation on your intended security goals. We have found it is most useful to provide guidance for how you rate the severity of reports (e.g., do you consider post-auth SQLi high or critical? are buffer overflows without demonstrated exploits capped at high? when is stored XSS medium, high, and critical?) This file can also state what the project does, where untrusted input enters, which components matter and which are out of scope, or anything else you think important or useful.

Two more files (optional but strongly recommended) let you say how you would like our reports and patches to look:

  • report_guidelines.md is for the format of each report: its layout and sections, length, tone, and which facts to state.
  • patch_guidelines.md is for the style of each patch: for example, how many comments to write, whether a fix should be very narrow or more general, which tests to add, and which formatter to use.

Start from templates/report_guidelines.md and templates/patch_guidelines.md. Both are plain text of at most 16 KiB.

These files only change how reports and patches are written; what is in scope and how you rate severity belong in threat_model.md. Each file is checked automatically before it is used. One that tries to change more than that, for example to leave findings out or to run unrelated commands, is ignored as a whole.

We suggest you run two commands before opening a pull request:

  • tools/validate.py checks projects/<name>/ against the rules above.
  • tools/check <name> builds your project the way our scanner will, and opens a shell in the finished image with no network. If your tests pass in this container, our scanner will likely work with your project. tools/check --qemu <name> does the same inside virtual machines laid out like the scanner's.

These tools need your host to have installed git, Docker, and Python 3 with PyYAML (pip install pyyaml); --qemu needs Linux on x86-64 with QEMU instead of Docker.

What happens after the merge

  1. The scanner imports the project, builds it online in an isolated VM, then moves the VM to a network with no Internet access for scanning. If the build fails, we will email primary_contact with an error message.
  2. We scan the project for vulnerabilities.
  3. Findings are emailed to primary_contact and any additional CCs with reproduction steps and a proposed patch where available.

You can edit your project at any time with a PR. To unenroll your project, set the disabled field to true to pause reports or remove projects/\<name\>/ to withdraw the project entirely.

Security considerations

Download Tool

Most Popular

View all →

Discover the most used tools by our community.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 The ultimate OSINT toolkit: investigation graph, confidence scoring, alias families, map, timeline and reports. 49 modules, 400+ sites, no API keys required.

    867
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock One-Tap Execution App (CVE-2026-43499)

    544
  • 3

    ARTEX

    baarai/artex

    LLM multi-agent autonomous penetration testing platform with planner/worker agents, asset and exploration graphs, MITM traffic recording, and vulnerability retesting.

    385
  • 4

    redforge

    cybersecurity91/redforge

    Browser-side red team toolkit with 102 offline tools for recon, payload generation, credential attacks, crypto, encoding and engagement reporting.

    360
  • 5

    looking-glass

    anonghosty/looking-glass

    CSIRT platform that ingests Shadowserver reports into MongoDB and replays them as interactive maps, analysis, findings, advisories and executive dashboards.

    302
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    The system of action for AI-native cybersecurity—where intent becomes governed execution, evidence becomes operational memory, and every operation improves the next.

    279

Explore all tools

Browse our collection of tools

View all tools →