Security●●●Banger
AegisBPF – Deterministic Runtime Enforcement via eBPF LSM
LSM hooks block operations synchronously; most eBPF security tools only alert asynchronously.
WizardryNiche GemBig Brain
erenari
203mo ago
Back to browse
GitHub Repository
Runtime visibility for Python MCP servers. Captures tool calls, session lifecycle, module imports (SHA-256), and subprocess execution as structured NDJSON. No code changes.
32 starsPython
MCP Python Hooks – sys.addaudithook and import hooks for MCP
by BlueRock-Jake·May 6, 2026·4 points·0 comments
AI Analysis
●●●BangerWizardryBig Brain
Zero-code runtime visibility for MCP servers using Python audit hooks is genuinely clever.
Strengths
- •Leverages sys.addaudithook to monitor transitive dependencies without code changes.
- •SHA-256 verification of every imported module prevents supply chain tampering.
- •Native support for MCP protocol events like session creation and tool calls.
Weaknesses
- •Audit hooks can introduce performance overhead in latency-sensitive applications.
- •Linux and macOS only; Windows support is missing from the current release.
Category
Target Audience
AppSec engineers and Python developers
Similar To
OpenTelemetry · Sysdig · Falco
Similar Projects
Security●●Solid
Sentinel – Open-source MCP security scanner (config, probe, container)
26 MCP-specific checks with GitHub Actions + SARIF, but confined to emerging protocol ecosystem.
Solve My ProblemNiche Gem
Siri_D
103mo ago
Security●●●Banger
VellaVeto – Fail-closed runtime proxy for MCP tool calls, in Rust
Formal verification + 20 threat layers for AI agent security at sub-5ms latency.
WizardryBig BrainShip It
paolovella
113mo ago
Security●●●Banger
Lilith-zero update: fast Rust security for agents
Transport-layer security for MCP agents stops exfiltration unlike prompt-based guardrails.
Solve My ProblemWizardry
gregojaca
201mo ago
Security●●●Banger
Sigil – source code security analysis for MCP servers (open source)
Source-code MCP security auditing. Existing scanners check descriptions; sigil reads actual code.
Zero to OneWizardrySolve My Problem
sigildev
203mo ago
Developer Tools●●●Banger
IronCurtain: A secure* runtime for AI agent loops
Sandbox agents via natural-language policy, not ambient authority—genuinely novel approach.
Big BrainZero to OneSolve My Problem
nielsprovos
113mo ago