2 KiB
2 KiB
Credits and references
Runtime dependencies
| Project | Use | License |
|---|---|---|
| Glaze, v7.5.0 | C++23 JSON and JSON-RPC encoding | MIT |
| libseccomp | Linux seccomp-bpf filter construction | LGPL-2.1 |
Glaze is fetched by CMake and linked privately. libseccomp is a Linux system dependency linked privately by the container module. Their licenses are compatible with GPLv3; they remain governed by their own terms.
Platform and protocol references
- Bubblewrap informed the Linux
user-namespace, UID/GID mapping, private-mount propagation,
pivot_root, and detached-old-root sequence. Glove neither links to nor copies Bubblewrap. - Model Context Protocol defines the JSON-RPC tool protocol implemented by the MCP module.
- Apple's Sandbox Profile Language is used for the macOS deny-default process profile. The underlying API is deprecated and may require replacement if removed by Apple.
- Yams is the external MCP server used by integration tests.
Research context
The following work informs the threat model and evaluation backlog:
- Debenedetti et al., “CaMeL: Defeating Prompt Injections by Design,” 2025.
- Wu et al., “IsolateGPT: An Execution Isolation Architecture for LLM-Based Agentic Systems,” 2024.
- Ruan et al., “Toolemu: Identifying the Risks of LM Agents with an LM-Emulated Sandbox,” ICLR 2024.
- Debenedetti et al., “AgentDojo: A Dynamic Environment to Evaluate Prompt-Injection Attacks and Defenses for LLM Agents,” NeurIPS 2024.
- Zhan et al., “InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents,” ACL 2024.
These works are references, not bundled code or evidence that Glove provides the same guarantees.
License
Glove is licensed under GPL-3.0-only.