lissy93/bug-bounties is a community-maintained directory, free API, and MCP server cataloguing 3,000+ active bug bounty and vulnerability disclosure programs for authorized research.
| Tool | lissy93/bug-bounties — community-maintained directory, MCP server and API of 3,000+ active bug bounty and vulnerability disclosure programs |
| Category | Reconnaissance data / program intelligence |
| Primary Use | Finding and filtering in-scope, authorized targets across 3,000+ bug bounty and VD programs via a free API or MCP integration |
| Safe Use | Intended for security professionals who need to confirm which assets are explicitly in-scope before conducting authorized testing or responsible disclosure |
| Telemetry Note | Purely an information resource — querying the directory leaves no trace on any target; defenders can use the same data to verify whether their own programs and scope definitions are accurately listed |
One of the most persistent frictions in authorized security research is simply knowing where the boundaries are. lissy93/bug-bounties attacks that problem from the data side: it is a community-maintained directory of more than 3,000 active bug bounty and vulnerability disclosure programs, exposed not just as a browsable listing but as a free API and an MCP (Model Context Protocol) server. Written in TypeScript and released under the MIT license, the project sits at the intersection of classic recon tooling and the newer wave of AI-assisted security workflows, which its ai-security and mcp topics make explicit. At roughly 431 stars it is a modest but actively positioned project from a well-known maintainer, and the directory-first architecture is what makes it interesting to examine.
The core value proposition is centralization. Rather than piecing together target lists from individual platform pages — HackerOne, Bugcrowd, Intigriti, standalone VD policies published on corporate sites — researchers can query one normalized dataset covering programs across all of those sources plus direct vulnerability disclosure arrangements. Each entry represents a program that is, at the time of curation, live and accepting reports. For a professional scoping an engagement or deciding where to focus weekend research, this collapses hours of manual discovery into a single queryable resource. The community-maintained aspect matters here: program scope drifts constantly, platforms delist targets, and a stale directory is worse than none, so the project's maintenance model is as important as its raw size.
Architecturally, three consumption surfaces stand out. The human-facing directory is the baseline; the API makes the dataset programmatically accessible, which means it can be wired directly into recon pipelines, target-management tooling, or scope-validation scripts; and the MCP server is the forward-looking piece. MCP is the emerging standard for letting LLM agents access external tools and data sources, so shipping an MCP interface means an AI assistant — a local agent, an IDE copilot, a custom automation — can pull live program data mid-workflow. Imagine an agent planning an assessment and being able to ask, structurally, whether a given asset is covered by an active program before touching it. That is the workflow this design anticipates.
The free-api topic signals the maintainers intend broad, unconstrained programmatic use, which is unusual and worth appreciating. Many comparable directories are scrape-gated or wrapped in paid tiers. For defenders and red-team tooling authors alike, a permissive, structured feed of disclosure programs is a building block: it can drive scope files, seed asset inventories, or serve as ground truth in reports. The TypeScript implementation also means the codebase is approachable for contributors who want to add or correct program entries, reinforcing the community-maintenance loop the project depends on.
The defensive angle deserves emphasis, because this is fundamentally a scope-authorization tool rather than an attack tool. The single most common way otherwise-legitimate researchers get into trouble is testing assets they believed were in scope and were not. A directory that answers 'is there an active program covering this organization, and what does it actually permit?' before any packet is sent is a compliance control, not an attack primitive. Security teams building internal research policies can use the same dataset in reverse: auditing whether their own organization's disclosure program is listed accurately, with correct contact channels and scope language.
The responsible-disclosure topic broadens the picture beyond classic paid bounty platforms. Vulnerability disclosure programs — the many organizations that accept reports without running a bounty — are heavily represented in a 3,000+ entry dataset, and for researchers these are often the more realistic targets: less competition, broader scope, and simpler processes. The directory's inclusion criteria, treating bounty programs and VD programs as one continuous spectrum of authorized reporting, reflects how the disclosure landscape actually works in practice.
Where this fits in an authorized workflow is straightforward. During the scoping phase of a bug bounty engagement or a permitted internal assessment, pulling the relevant program entry confirms what is in bounds, what is excluded, and how reports should be submitted. In automation, the API can gate recon tooling so that subdomain enumeration or content discovery only proceeds against assets verified as in-scope — a guardrail pattern more teams should adopt. And in the emerging agentic-security space, the MCP server makes that guardrail available to LLM-driven workflows natively, which is the project's most differentiated bet.
Caveats are the usual ones for community data: curation is only as current as its contributors, so any program entry should be verified against the authoritative source — the platform page or the organization's own security.txt and policy documents — before acting on it. Treat the directory as a discovery and prioritization layer, not the final word on scope. With that discipline, lissy93/bug-bounties is a genuinely useful piece of civic security infrastructure: free, permissively licensed, community-scrubbed, and designed for exactly the boundary-checking that separates authorized research from trespass.
lissy93/bug-bounties.Educational analysis for authorized security professionals. Use only in controlled, authorized environments.
Related coverage
0 comentários:
Post a Comment
Note: Only a member of this blog may post a comment.