kingfisher

Detect and Validate Secrets Anywhere. Map Access. Revoke Fast.

Find leaked secrets. Validate what’s live. Map the blast radius. Revoke fast.

Kingfisher is a fast, free, open-source secret scanner and credential response tool built in Rust. It detects leaked secrets with hundreds of built-in rules, validates which credentials are active, maps access for supported credential types, and revokes supported secrets—all in one Apache-2.0-licensed release:

Detect → Validate → Map → Triage → Revoke

• scan source code, Git history, cloud storage, container images, archives, and developer platforms
• validate which credentials are active
• map a leaked secret’s identity, permissions, and reachable resources
• triage findings in a browser, or output as SARIF, JSON, TOON
• send real-time alerts to Slack, Microsoft Teams, Discord, Mattermost, Google Chat, or any JSON webhook
• revoke supported secrets

Defender workflow: Follow the end-to-end defender workflow for secret detection, validation, notifications, blast-radius mapping, and revocation.

Embed Kingfisher in Rust, Python, or JavaScript

Add Kingfisher’s fast secret scanning, live credential validation, and revocation to your own applications through Python, Rust, or the native Node SDK preview. Everything runs in-process, with no CLI subprocess required.

• Python: Install uv add kingfisher-secret-scanner and import kingfisher_sdk. See the Python SDK guide and runnable examples for scanning, validation, rule inspection, and explicit revocation.
• Rust: Use kingfisher-scanner. See the library guide and runnable examples for scanning and validation, plus revocation via kingfisher-scanner.
• JavaScript / TypeScript: The kingfisher-secret-scanner npm SDK 0.1.0 is a preview of npm support, wrapping the same Rust crates with napi-rs. See the Node SDK guide and source/examples. It targets Node.js, including VS Code workspace extensions, rather than browser/WASM runtimes. Start with the runnable recipe walkthrough; maintainers can follow the npm trusted-publishing setup.

Scan Targets

Kingfisher handles local files and directories, Git repositories and history, compressed and office-document archives, SQLite databases, Python bytecode, Docker images, source-hosting organizations, cloud object storage, collaboration tools, and API-development platforms.

Files / Dirs	Local Git	GitHub	GitLab	Azure Repos	Bitbucket	Gitea	Hugging Face
Files / Dirs	Local Git	GitHub	GitLab	Azure Repos	Bitbucket	Gitea	Hugging Face

Docker	Jira	Confluence	Slack	Teams	Postman	AWS S3	Google Cloud	Azure Blob Storage
Docker	Jira	Confluence	Slack	Teams	Postman	AWS S3	Cloud Storage	Azure Blob Storage

For target-specific commands, authentication, scope, and pagination behavior, use the platform integration guide.

Discover, validate, and revoke — open source options

All four tools detect secrets. Kingfisher and Betterleaks 2.0 RC5 also include credential revocation for supported providers.

✅ Built in · ❌ Not supported as a built-in capability.

Built for Speed and Accuracy

Kingfisher’s multithreaded Vectorscan engine recorded the lowest runtime on every repository in the published benchmark suite, from small projects through the Linux kernel and GitLab monorepo. Lower runtimes are better.

See the benchmarks for validation results, network-request counts, and the test environment, and deployment options for local, CI, and embedded use.

Why Kingfisher

Live validation, advanced cloud blast-radius analysis, visual triage, real-time alerts, and supported revocation all ship in the free, open source Kingfisher. There is no separate paid or enterprise tier.

Examples for different targets (replace credential placeholders with values from your secret store):

# Local files or a repository, including its Git history
kingfisher scan /path/to/code --format json

# Working tree only, without live credential checks
kingfisher scan /path/to/code \
    --git-history none \
    --no-validate \
    --format json

# Staged changes, archives, or office documents
kingfisher scan . \
    --staged \
    --format json
kingfisher scan backup.tar.gz credentials.xlsx --format json

# Remote Git repository; private GitHub repositories use KF_GITHUB_TOKEN
KF_GITHUB_TOKEN="..." \
    kingfisher scan https://github.com/my-org/my-repo.git --format json

# GitHub organization, or a user's repositories and gist history
KF_GITHUB_TOKEN="..." \
    kingfisher scan github \
        --organization my-org \
        --format json
KF_GITHUB_TOKEN="..." \
    kingfisher scan github \
        --user alice \
        --include-gists \
        --format json

# GitLab group, including subgroups and snippets
KF_GITLAB_TOKEN="..." \
    kingfisher scan gitlab \
        --group my-group \
        --include-subgroups \
        --include-snippets \
        --format json

# Azure DevOps repositories
KF_AZURE_TOKEN="..." \
    kingfisher scan azure \
        --azure-organization my-org \
        --format json

# Bitbucket workspace and snippets
KF_BITBUCKET_TOKEN="..." \
    kingfisher scan bitbucket \
        --workspace my-team \
        --include-snippets \
        --format json

# Gitea organization; private clones may also need KF_GITEA_USERNAME
KF_GITEA_TOKEN="..." KF_GITEA_USERNAME="scanner" \
    kingfisher scan gitea \
        --organization my-org \
        --format json

# Hugging Face repositories and buckets visible to an organization
KF_HUGGINGFACE_TOKEN="..." \
    kingfisher scan huggingface \
        --huggingface-organization my-org \
        --format json

# Hugging Face bucket: selected paths, with reusable content and listing checkpoints
KF_HUGGINGFACE_TOKEN="..." \
    kingfisher scan huggingface \
        --huggingface-bucket hf://buckets/my-org/data \
        --path-prefix configs/ \
        --file-extension env \
        --object-cache-dir ./object-cache \
        --format json

# AWS S3 using a local profile (public buckets can be scanned without credentials)
kingfisher scan s3 my-bucket \
    --profile security \
    --path-prefix configs/ \
    --format json

# AWS S3 using explicit credentials; add KF_AWS_SESSION_TOKEN for temporary credentials
KF_AWS_KEY="..." KF_AWS_SECRET="..." \
    kingfisher scan s3 my-bucket --format json

# Google Cloud Storage using a service-account file (also supports default credentials)
GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json" \
    kingfisher scan gcs my-bucket --format json

# Azure Blob Storage: container URL and a SAS token with read/list permission
KF_AZURE_STORAGE_SAS="sv=...&sp=rl&sig=..." \
    kingfisher scan azure-blob https://myaccount.blob.core.windows.net/my-container \
        --path-prefix configs/ \
        --skip-path-prefix configs/generated/ \
        --file-extension env \
        --object-cache-dir ./object-cache \
        --format json

# Container image from a registry, or a saved image archive; private registries use KF_DOCKER_TOKEN="user:password"
kingfisher scan docker ghcr.io/owasp/wrongsecrets/wrongsecrets-master:latest-master --format json
kingfisher scan docker \
    --archive image.tar \
    --format json

# Jira Cloud issues and comments; omit KF_JIRA_USER for a server bearer token
KF_JIRA_USER="[email protected]" KF_JIRA_TOKEN="..." \
    kingfisher scan jira \
        --url https://example.atlassian.net \
        --jql "updated >= -30d" \
        --skip-project ARCHIVE \
        --include-comments \
        --format json

# Confluence Cloud pages; omit KF_CONFLUENCE_USER for a server bearer token
KF_CONFLUENCE_USER="[email protected]" KF_CONFLUENCE_TOKEN="..." \
    kingfisher scan confluence \
        --url https://example.atlassian.net/wiki \
        --cql 'space = "ENG"' \
        --format json

# Slack messages and files, or Microsoft Teams messages
KF_SLACK_TOKEN="..." \
    kingfisher scan slack "api_key OR password" --format json
KF_TEAMS_TOKEN="..." \
    kingfisher scan teams "api_key OR password" --format json

# Postman collections and environments in a workspace
KF_POSTMAN_TOKEN="..." \
    kingfisher scan postman \
        --workspace my-workspace-id \
        --format json

See usage for the full command reference and integrations for platform-specific examples, including Azure Blob Storage, GitLab, Azure DevOps, Gitea, Slack, and Jira.

See Kingfisher in action with basic scan and validation examples, the end-to-end defender workflow, platform-specific scan examples, blast-radius examples, direct revocation, and CI and pre-commit deployment.

Real-Time Alerts

Send findings to the place your responders already watch. When a scan completes, Kingfisher can POST a summary with active/inactive counts, top rules, optional per-finding details, and a link to the full report.

# Repeat --alert-webhook to fan out; Slack and Teams are inferred from the URL hosts
kingfisher scan /path/to/repository \
    --alert-webhook "$SLACK_SECURITY_WEBHOOK" \
    --alert-webhook "$TEAMS_SECURITY_WEBHOOK"

Slack, Microsoft Teams, Discord, Mattermost, Google Chat, and generic JSON webhooks are supported. Confidence, validation, and access-map filters plus detail levels can be set globally from the CLI or per destination in kingfisher.yaml; payloads also carry stable finding fingerprints for deduplication. Alert delivery is best-effort and never changes the scan exit code.

See the alert webhook guide for payload formats, --alert-* flags, finding filters, and configuration examples.

Map the Blast Radius. Revoke the Credential.

For supported credentials, --blast-radius goes beyond a live/inactive verdict. It maps the credential identity, recorded permissions, candidate role or service-account paths, and resource scopes. This includes AWS role-assumption and GCP service-account impersonation analysis, bounded Bedrock model discovery and availability checks, and exact read methods accepted by a bounded Google API-key probe allowlist.

The HTML triager opens directly into an investigation workspace. Review active credentials and coverage gaps, then trace each identity to policy evidence, candidate privilege paths, and recorded resources. AWS, GCP, and Azure mappings are included in the open-source release.

A 10-second walkthrough of actual viewer screens using synthetic data. Prefer a still image? Open the policy investigation or resource map.

The Blast Radius view separates observed API access from policy grants and conditional paths. Search evidence, filter providers, or switch to the resource map. Risk explanations and coverage gaps are included automatically using the same API results. Use --aws-region or --gcp-project to expand discovery. Existing report formats and legacy severity fields remain supported.

Screenshots use a synthetic, credential-free report. See the triager workflow and collection limits.

Kingfisher also provides explicit, defender-led revocation for supported credentials. Revocation is opt-in and is exposed only where Kingfisher has a bounded provider workflow; responders should always confirm the target and operational impact before containment.

Performance, Accuracy, and Extensible Rules

• Performance: multithreaded, Hyperscan‑powered scanning built for huge codebases
• Extensible rules: Betterleaks is the main catalog, with selected Veles detectors and rare documented first-party rules filling gaps; both Betterleaks TOML (.toml) and the Kingfisher rule format (.yml/.yaml) are fully supported via --rules-path (built-in rules, docs/RULES.md)
• Validation and defender-led revocation: validate discovered credentials live, then revoke supported credentials from the CLI. For supported provider flows, responders can contain a leaked token even when its owner is unknown or has left the company (docs/USAGE.md, docs/REVOCATION_PROVIDERS.md)
• Blast-radius mapping in the open-source release: use --blast-radius (alias --access-map) to map supported credentials to identities, permissions, candidate role/service-account paths, and resource scopes. All 43 providers—including AWS IAM, GCP hierarchy/impersonation, and Azure Graph/ARM RBAC mapping—are included in the Apache-2.0 release (blast-radius docs)
• Broad provider coverage: detect and validate credentials across cloud, AI, developer tooling, databases, SaaS, messaging, identity, and cryptographic systems through the Betterleaks- and Veles-based candidate catalog
• Compressed Files: Supports extracting and scanning compressed files for secrets, including tar.gz/bz2/xz, ZIP-family containers (zip, jar, docx, xlsx, pptx, odt, epub, hwpx, and more), asar, HWP (Hancom OLE2/CFBF binary with DEFLATE/zlib stream decoding), and EGG (ALZip; raw-byte scanning)
• SQLite Database Scanning: Automatically extracts and scans SQLite database contents for secrets stored in table rows
• PDF Text Scanning: Extracts page text from PDFs, including Git history and archive members, with decompression/output limits; no OCR or password-protected PDF support
• Python Bytecode (.pyc) Scanning: Extracts and scans string constants from compiled Python (.pyc, .pyo) files
• Baseline management: generate and track baselines to suppress known secrets (docs/BASELINE.md)
• Checksum-aware custom detection: Kingfisher custom rules can verify token checksums offline before validation (checksum intelligence)
• Report Viewer (local + hosted): Visualize and triage Kingfisher, SARIF, Gitleaks, and TruffleHog output locally with kingfisher view ./report.json or online with the hosted viewer. Multiple files, directories, and imported third-party reports are merged and deduplicated. See docs/USAGE.md.
• Audit reporting: Generate compliance-oriented HTML reports with scan metadata and validation ordering
• Embedding: Scan in-process from Rust, Python, or the Node SDK preview, with explicit validation and revocation APIs

Basic Usage Demo

kingfisher scan /path/to/scan --view-report

NOTE: Replay has been slowed down for demo

Getting Started

Quick Start

Install with your preferred package manager:

# Homebrew (macOS/Linux)
brew install kingfisher

# PyPI wrapper
uv tool install kingfisher-bin

See the full installation guide for all installation methods, installer options, and pre-commit hook installation.

The examples below use Bash syntax. Replace credential placeholders with values from your secret store; local scans need no platform credentials. See the environment-variable reference for other authentication options.

Then scan a repository, including its Git history:

kingfisher scan /path/to/repository --format json

Open the results in the bundled local viewer:

kingfisher scan /path/to/repository --view-report

Scan files and Git history

# Clone a private GitHub repository and scan its Git history (omit the token for public repositories)
KF_GITHUB_TOKEN="..." \
    kingfisher scan https://github.com/my-org/my-repo.git --format json

# Scan checked-out files without Git history or live validation
kingfisher scan /path/to/repository \
    --git-history none \
    --no-validate \
    --format json

# Check staged changes before committing
kingfisher scan . \
    --staged \
    --format json

# Scan an archive or an office document
kingfisher scan backup.tar.gz credentials.xlsx --format json

# Focus on configuration files and skip generated content
kingfisher scan /path/to/repository \
    --path-prefix configs/ \
    --skip-path-prefix configs/generated/ \
    --file-extension env \
    --file-extension yaml \
    --format json

# Show only live credentials and map their access
kingfisher scan /path/to/repository \
    --only-valid \
    --blast-radius \
    --format json

# Include active credentials and high-confidence assumed or locally derived secrets
kingfisher scan /path/to/repository \
    --validation-filter actionable \
    --format json

Save and combine reports

# Save a JSON report and open it later
kingfisher scan /path/to/repository \
    --format json \
    --output findings.json
kingfisher view findings.json

# Export SARIF for code-scanning integrations
kingfisher scan /path/to/repository \
    --format sarif \
    --output findings.sarif

# Generate a standalone HTML audit report
kingfisher scan /path/to/repository \
    --format html \
    --output audit.html

# Import and combine Kingfisher, Gitleaks, and TruffleHog reports
kingfisher view findings.json gitleaks.json trufflehog.jsonl

# Load every supported report in a directory
kingfisher view ./reports/

Scan source-hosting organizations

Use a token with read access to the repositories you want to scan. See the integration guide for provider-specific permissions and self-hosted endpoints.

# GitHub organization
KF_GITHUB_TOKEN="..." \
    kingfisher scan github \
        --organization my-org \
        --format json
# GitHub user and their gist history
KF_GITHUB_TOKEN="..." \
    kingfisher scan github \
        --user alice \
        --include-gists \
        --format json

# GitLab group, including nested subgroups
KF_GITLAB_TOKEN="..." \
    kingfisher scan gitlab \
        --group my-group \
        --include-subgroups \
        --include-snippets \
        --format json

# Azure Repos organization
KF_AZURE_TOKEN="..." \
    kingfisher scan azure \
        --azure-organization my-org \
        --format json

# Bitbucket workspace
KF_BITBUCKET_TOKEN="..." \
    kingfisher scan bitbucket \
        --workspace my-team \
        --include-snippets \
        --format json

# Gitea organization
KF_GITEA_TOKEN="..." KF_GITEA_USERNAME="scanner" \
    kingfisher scan gitea \
        --organization my-org \
        --format json

# Hugging Face organization
KF_HUGGINGFACE_TOKEN="..." \
    kingfisher scan huggingface \
        --huggingface-organization my-org \
        --format json

# Preview the GitHub repository scope without scanning
KF_GITHUB_TOKEN="..." \
    kingfisher scan github \
        --organization my-org \
        --list-only

Scan cloud storage

For S3 and GCS, use a bucket name without s3:// or gs://. Azure Blob uses a full container URL. Use --path-prefix to select object names; S3 and GCS also accept the existing --prefix option. Omit prefix selection to scan the whole bucket or container.

# AWS S3: whole bucket with explicit credentials
KF_AWS_KEY="..." KF_AWS_SECRET="..." \
    kingfisher scan s3 my-bucket --format json

# AWS S3: temporary credentials also require the session token
KF_AWS_KEY="..." KF_AWS_SECRET="..." KF_AWS_SESSION_TOKEN="..." \
    kingfisher scan s3 my-bucket --format json

# AWS S3: only objects under backups/production/, using a named AWS profile
kingfisher scan s3 my-bucket \
    --prefix backups/production/ \
    --profile security \
    --format json

# Google Cloud Storage: whole bucket using a service-account file
GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json" \
    kingfisher scan gcs my-bucket --format json

# Google Cloud Storage: only objects under exports/daily/, using existing Application Default Credentials
kingfisher scan gcs my-bucket \
    --prefix exports/daily/ \
    --format json

# Azure Blob Storage: container SAS with read (r) and list (l) permissions
KF_AZURE_STORAGE_SAS="sv=...&sp=rl&sig=..." \
    kingfisher scan azure-blob \
        https://myaccount.blob.core.windows.net/my-container --format json

# Azure Blob Storage: account-key authentication (account name is inferred for standard Azure endpoints)
KF_AZURE_STORAGE_KEY="..." \
    kingfisher scan azure-blob https://myaccount.blob.core.windows.net/my-container --format json

# Azure Blob Storage: an existing Entra access token for https://storage.azure.com/
KF_AZURE_STORAGE_TOKEN="..." \
    kingfisher scan azure-blob https://myaccount.blob.core.windows.net/my-container --format json

# Hugging Face bucket: selected paths with resumable downloads
KF_HUGGINGFACE_TOKEN="..." \
    kingfisher scan huggingface \
        --huggingface-bucket hf://buckets/my-org/data \
        --path-prefix configs/ \
        --object-cache-dir ./object-cache \
        --format json

Authenticate to S3 with KF_AWS_KEY / KF_AWS_SECRET or --profile; GCS uses Application Default Credentials, or an explicit --service-account /path/to/key.json. See the S3 and GCS guides for authentication and public-bucket examples. Azure Blob accepts a container URL; set exactly one of the SAS, account-key, or Entra-token variables, or omit all three for anonymous access to a public container with list/read permissions; see the Azure Blob guide.

Use --path-prefix, --skip-path-prefix, --file-extension, and --skip-file-extension to select paths across local files, Git repositories, archive members, and cloud storage. S3, GCS, Azure Blob, and Hugging Face buckets apply selection before downloading and accept --object-cache-dir to resume interrupted listings and reuse unchanged content. Cached content is rescanned so each report stays complete. Custom HTTP verification services can use OAuth2 password-grant authentication.

Scan containers or run Kingfisher in Docker

# Scan a registry image
kingfisher scan docker ghcr.io/owasp/wrongsecrets/wrongsecrets-master:latest-master --format json

# Pull a private registry image
KF_DOCKER_TOKEN="user:password" \
    kingfisher scan docker ghcr.io/my-org/my-image:latest --format json

# Scan an image exported with docker save
kingfisher scan docker \
    --archive image.tar \
    --format json

# Run Kingfisher against the current directory without installing it
docker run \
    --rm \
    -v "$PWD":/src \
    ghcr.io/mongodb/kingfisher:latest scan /src \
    --format json

# Run in Docker and serve the report viewer to the host
docker run \
    --rm \
    -v "$PWD":/src \
    -p 127.0.0.1:7890:7890 \
    ghcr.io/mongodb/kingfisher:latest scan /src \
    --view-report \
    --view-report-address 0.0.0.0

For the Docker viewer, open localhost:7890 in your browser.

Scan issues, documents, and messages

Jira and Confluence Cloud API tokens require the account email as well as the token. For server bearer tokens, omit KF_JIRA_USER or KF_CONFLUENCE_USER. See the integration guide for required scopes and search behavior.

# Jira Cloud: search a project and include issue comments
KF_JIRA_USER="[email protected]" KF_JIRA_TOKEN="..." \
    kingfisher scan jira \
        --url https://example.atlassian.net \
        --jql "project = SEC" \
        --include-comments \
        --format json

# Confluence Cloud: scan pages in a space
KF_CONFLUENCE_USER="[email protected]" KF_CONFLUENCE_TOKEN="..." \
    kingfisher scan confluence \
        --url https://example.atlassian.net/wiki \
        --cql 'space = "ENG"' \
        --format json

# Slack: search messages and associated files
KF_SLACK_TOKEN="..." \
    kingfisher scan slack "api_key OR password" --format json

# Microsoft Teams: search messages
KF_TEAMS_TOKEN="..." \
    kingfisher scan teams "api_key OR password" --format json

# Postman: scan a workspace's collections and environments
KF_POSTMAN_TOKEN="..." \
    kingfisher scan postman \
        --workspace my-workspace-id \
        --format json

Validate or revoke a known credential

# Validate a GitHub personal access token directly
export GITHUB_PAT="..."
kingfisher validate \
    --rule github-pat "$GITHUB_PAT" \
    --format json

# Revoke that token when ready to contain the exposure
kingfisher revoke \
    --rule github-pat "$GITHUB_PAT" \
    --format json

See direct validation and supported revocation providers for other credential types.

See the installation guide for pre-built binaries, Docker, mise, Windows, pre-commit hooks, release verification, and source builds. See the usage guide for validation filters, output formats, scan scope, and command examples.

Live validation and blast-radius mapping make authorized requests to provider APIs. Review the relevant documentation and use them only where you are authorized to inspect the target account.

Common Workflows

Output for People and Machines

Kingfisher supports human-readable output plus TOON, JSON, JSONL, SARIF, BSON, and HTML reports. Use JSON for structured output in scripts and integrations:

kingfisher scan /path/to/repository \
    --format json \
    --no-update-check

Machine consumers should use structured validation outcomes and finding fingerprints rather than parsing display labels. See output and validation semantics, finding fingerprints, and the full documentation index.

Rust Embedding Examples

The library guide includes runnable examples for kingfisher-core, kingfisher-rules, kingfisher-scanner, and the kingfisher-bin library. Copyable integration recipes cover file batches, private rules, async services, and scanning plus HTTP validation. For example, from a repository checkout:

cargo run --locked -p kingfisher-scanner --example http_validation --features validation

Verifying Releases

Before installing a pinned release, verify that your download came from Kingfisher’s release workflow and matches the version you requested. This catches tampered downloads and substitution with an older genuine release, which matters when Kingfisher is your CI secrets gate. Follow the release verification steps.

Documentation

• Documentation index: task-oriented map of every user, operator, rule-author, and developer guide in the repository.
• Hosted documentation: searchable rendered version.
• llms.txt: compact machine-readable map of the documentation set.
• Cargo publishing: prepared rule bundles, provenance, and crate releases.
• Architecture: codebase layout and data flow.
• Changelog: release-by-release behavior changes.

Project

Kingfisher is used in MongoDB’s production security workflows and is integrated by projects such as Prowler, MegaLinter, Osmedeus, DefectDojo, and hk. Weblate uses its pre-commit hook. Read more about its lineage, evolution, and public adoption.

Contributions are welcome. See CONTRIBUTING.md, report vulnerabilities through SECURITY.md, and file feature requests in GitHub Issues.

Kingfisher is licensed under the Apache License 2.0.