MCP Tools DirectoryAbout this integration
Scan prompts, tool definitions and model output for injection, and guard agent tool calls.
- Transport
- stdio
- Authentication
- none
- Initial setup
- none
- Runtime
- unattended
- Evidence
- documented
- Version
- 3.1.1
- Package
- prompt-protection
- Last compatibility test
- Not independently tested
Connect your agent
prompt-protectionPublisher documentation and manifest were reviewed without executing the package. The listing describes documented behavior only and retains the existing published package version separately from the captured source-manifest version.
Capabilities: Scan prompts, tool definitions, and outputs, Register untrusted sources and check proposed tool calls, Spotlight text and detect canary leakage
Connected profiles
Additional details
io.github.mughalhere/prompt-protection
Source ↗ · Checked 2026-09-173.1.1
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17npm
Source ↗ · Checked 2026-09-17prompt-protection
Source ↗ · Checked 2026-09-194.0.0-rc.1
Source ↗ · Checked 2026-09-19Publisher documents an MCP executable that exposes seven local scanning, source-registration, tool-call, spotlighting, and canary tools.
Source ↗ · Checked 2026-09-19The core operates fully offline with optional networked model adapters disabled unless explicitly wired by the operator.
Source ↗ · Checked 2026-09-19Run npx -y prompt-protection-mcp as a local stdio MCP server; the optional MCP SDK peer is required for that surface.
Source ↗ · Checked 2026-09-19Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-19