v1.2.0: external agents, MCP over HTTP, secrets, and AI Act Art. 50 - #96
Merged
Merged
Conversation
Covers everything merged since 1.1.0: external agents behind an
OpenAI-compatible endpoint, MCP over HTTP, ${VAR} secrets, Art. 50
disclosure and image marking, plus the corrected compliance wording and
the README demo links.
The security section is separate rather than folded into Added, because
two of those entries describe credentials that used to be able to leave
the host and now cannot.
The announcement leads with the external-agent change rather than the
feature list. It is the one that answers "why would I rebuild my agents
to try your dashboard", which is the question the project has been
failing to answer for six months. The Art. 50 wording says explicitly
that MATE does not make anyone compliant.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01ToCUNy2SqwvfTq4a6xfSk1
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Release v1.2.0 adds support for external agents (pointing at any OpenAI-compatible endpoint), MCP over HTTP, environment variable secrets in config, and EU AI Act Article 50 compliance features (AI disclosure and generated image marking).
Changes
model_base_urlandmodel_api_key, applying RBAC, guardrails, evals, versioning, audit, and cost tracking without requiring agents to be built in MATEurlare reached directly over HTTP/SSE instead of spawningnpx mcp-remoteas a subprocess; stdio servers unchanged${VAR}secrets — MCP server entries and agent endpoint keys can reference server environment variables, keeping credentials out of the database and config exports; unset variables skip the server or refuse to build the agenttrainedAlgorithmicMediamarker ahead of the 2 December 2026 deadlineMCPToolsetrenamed toMcpToolset— ADK 2.3 keeps the old name as a deprecated subclass for backward compatibility${VAR}references refuse to build; stored endpoint keys never reach the browserdocuments/AI_ACT.mdexplaining what applies when and what MATE does and does not do about complianceNotes
The AI Act features are defaults that point the right way; they do not make anyone compliant. Compliance obligations sit with the provider or deployer of a system, not the framework. The audit trail and waiver mechanism ensure decisions and their justifications are recorded and cannot come apart.
The external agent path is the primary focus for feedback — testing against whatever endpoint you're running is most valuable.
https://claude.ai/code/session_01ToCUNy2SqwvfTq4a6xfSk1