MCP Tools DirectoryAbout this integration
AI agent security: prompt injection detection, semantic memory, output scanning, prompt hardening
- Transport
- stdio
- Authentication
- none
- Initial setup
- none
- Runtime
- unattended
- Evidence
- documented
- Version
- 0.2.2
- Package
- agent-immune
- Last compatibility test
- Not independently tested
Connect your agent
agent-immuneSecurity scores are one layer of defense and not a replacement for deterministic controls. The package and models were not loaded or executed.
Capabilities: prompt threat assessment, model-output scanning, adaptive threat memory
Connected profiles
Additional details
io.github.denial-web/agent-immune
Source ↗ · Checked 2026-09-170.2.2
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17pypi
Source ↗ · Checked 2026-09-17agent-immune
Source ↗ · Checked 2026-09-190.2.2
Source ↗ · Checked 2026-09-19The documented module command starts a local MCP server; stdio is the default desktop transport and loopback HTTP modes are optional.
Source ↗ · Checked 2026-09-19The local stdio server's documented install and launch path contains no service credential or external account requirement.
Source ↗ · Checked 2026-09-19The reviewed connection is python -m agent_immune serve --transport stdio, backed by the manifest-declared CLI artifact.
Source ↗ · Checked 2026-09-19Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-19