Grantex is an open-source delegated authorization protocol and reference implementation for AI agents. It gives each agent a verifiable identity and scoped, time-limited, revocable authority from a human or organization, using multi-agent delegation, service-side verification, and audit records.
The protocol is described as complementing OAuth 2.0 and MCP. In that model, OAuth handles application and user authorization, MCP connects models to tools, and Grantex proves which agent may perform which action for which principal. Its stated problem domain is AI agent authorization, where it replaces shared or all-or-nothing credentials with per-agent identity, least-privilege scopes, explicit approval, expiry, delegation constraints, revocation, and attributable audit records.
The page lists several implemented areas and packages. These include offline consent bundles and on-device verification examples for Gemma agents, a published OAuth endpoint package for single-process evaluation, DPDP Act 2023 and EU AI Act control mappings, structured consent records, purpose limitation, and audit-ready exports, a public trust registry with DID profiles, public keys, and DNS ownership verification, and anomaly detection based on four SQL-backed activity checks with lifecycle APIs and stored custom-rule and channel configuration. It also mentions a CLI command, grantex verify, for decoding claims locally, verifying signatures with JWKS, and optionally using a network-backed live revocation check.
Grantex packages are independently versioned. The page lists TypeScript, Python, and Go SDKs, along with MCP Auth and other capability areas. It also says that framework adapters are available for OpenAI Agents SDK, Anthropic, LangChain, CrewAI, Google ADK, Vercel AI, AutoGen, Express, FastAPI, and MCP. The protocol is available on GitHub, with docs, a playground, releases, examples, and a dashboard linked from the site.
Licensing and release information are stated on the page as Apache 2.0 and Protocol v1.0. It also notes published releases and current public releases, with verification against npm, PyPI, and the Go module proxy on 12 July 2026.
In the Auth & identity space, Grantex takes a focused approach. It provides secure, auditable delegated authorization and identity management for AI agents using open standards. Grantex is an open-source project aimed at ai infrastructure developers. The project is open source (Apache-2.0). It runs on the web, the command line, embeddable surfaces, and API, and it can be self-hosted.
Behind Grantex is Grantex, and it first shipped in 2026. Development happens publicly on GitHub with 27 stars and 1k commits in the last 90 days. Among its 5 catalogued features are OAuth 2.0 protocol, agent identity, and verifiable credentials. It exposes integrations via an MCP server and a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match