security-manager — Claude Code subagent for consensus security
security-manager — a Claude Code subagent that embeds cryptographic security into distributed consensus protocols. Runs on critical operations, verifying threshold signatures, detecting attacks, and encrypting node communications.
no license ★ 850 updated 2026-10-26
.claude/agents/security-manager.mdWhat security-manager does
Deploys threshold cryptography, zero-knowledge proofs, and attack detection over consensus protocols. Manages distributed key generation, rotation, TLS 1.3 encryption, and message authentication between nodes. Includes detectors for Byzantine, Sybil, and Eclipse attacks with real-time triggers.
Implemented as pre/post hooks: before a task, activates cryptographic verification; after, runs a security audit. Works with secp256k1 elliptic curves, Lagrange interpolation for combining partial signatures, and Bulletproof range proofs. Without an elliptic curve cryptography library, the subagent will not start.
What it helps with
- verify threshold signatures in a distributed system
- detect consensus attacks (Byzantine, Sybil, Eclipse)
- generate distributed keys without a trusted party
- prove knowledge of discrete logarithm without revealing the secret
- encrypt node communications with message authentication
Example prompts for the agent
- Run distributed key generation for three nodes with a two-signature threshold
- Create a threshold signature of a message from five participants
- Verify a zero-knowledge proof for a node's public key
How to install security-manager
Manually: put the files in place
.claude/agents/security-manager.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/security-manager.tomlCodex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/
Manually: put the files in place
.cursor/agents/security-manager.mdCursor also reads .claude/agents/
Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean
Questions about security-manager
How do I install security-manager in Claude Code?
Save the subagent file to .claude/agents/security-manager.md at the project root. Claude Code will pick it up automatically on the next task run.
What does security-manager do?
Embeds cryptographic security into distributed consensus protocols: generates threshold signatures, detects consensus attacks, and encrypts node communications. Runs via pre and post hooks around task execution.
Which attacks does security-manager detect?
Byzantine attacks (malicious nodes), Sybil attacks (multiple fake identities), Eclipse attacks (node isolation from the network), and DoS attacks on consensus protocols. Triggers in real time when anomalies are detected.