Security plugins

Разбор крипто-адресов Tron и EVM только на чтение: чей адрес, куда ушли деньги, что на нём лежит, связи, отравление адресов и разрешения на списание. crypttrace плюс бесплатные API Etherscan, TronGrid, Tronscan, Blockscout.

1 skill

Run /codesoteria:scan <path> to perform a comprehensive security audit

Solana token safety scanner — $0.02/check via x402 USDC, no API keys, no accounts

Run NebuSec code and cloud security workflows from Claude Code.

Interactive wizard that builds a harness in your project (settings.json, CLAUDE.md, hooks, custom commands) with confirm-before-write steps and resume support

Automated Cross-Account Data Isolation security auditing for multi-tenant codebases. Auto-detects tech stack and tenancy configuration, uses LSP-based call graph analysis, generates exploit scripts.

Captures complete coding-agent sessions — thinking, tool calls and results, the span DAG and subagent topology — and enforces policy at PreToolUse. Capture is on by default; enforcement ships off and is enabled per rule.

AI-powered security review via /security-review command (Anthropic)

Checks open source projects for security leaks, privacy issues, and license compliance before release

Ship secure code at agent speed. Clover silently reviews every plan, surfaces missing security requirements, and folds them into the work before a single line is written — no questions, no detours.

SAST analysis, dependency vulnerability scanning, OWASP Top 10 compliance, container security scanning, and automated security hardening

SOC2, HIPAA, and GDPR compliance validation, secrets scanning, compliance checklists, and regulatory documentation

API security hardening, authentication implementation, authorization patterns, rate limiting, and input validation

Deterministic guardrails for nondeterministic agents: tested, configurable hooks for AI coding agents.

Blocks shell commands that wipe your home folder, the project, disks, protected branches or databases, and snapshots uncommitted work to a git ref first.

Give Claude a password through a masked field; it goes to the macOS Keychain, never to the model.

入れたほかの mod の通信・コマンド・ファイル操作を、実行前に止めて確認し、mod ごとに許すかを決める

変更中のファイルを Semgrep で検査し、検出結果を画面の右で1件ずつ選別、本物だけ Claude に修正を依頼

Turns permanent deletes into the system trash and blocks no-way-back commands (git reset --hard, push --force, format) until the user says yes

Before git push or gh publishing, scans the commits about to leave for keys, tokens and passwords (gitleaks, with fallback rules) and blocks the push on a hit

Explains each Claude Code sandbox block in plain words and turns recurring blocks into reviewed, one-press, undoable edits of your user settings.

After a hook, deny rule, or auto mode blocks a tool call, Arrester stops later calls that reach the same result another way or read or change your guards. After your No, it stops only exact retries.

Denies .env reads, git force pushes and rm -rf on / or your home folder. Pattern-based.

Hand the agent secrets it can use in Bash as $NAME without ever seeing their values

Ask for credentials in a form above the prompt, keep them in the keychain, run commands with them masked.

Holds risky shell commands (rm -rf, git reset --hard, git clean, force push, migrations) and shows what they would change, with Proceed and Cancel buttons. By Anthropic.

Stops secrets from leaking before they are written, printed or committed to git: masks prompts and tool output, holds or blocks commands that contain secrets. Reviewed: Caution.

코드, 도구 출력, 대화 속 키와 토큰을 모델에게서 가리고, 도구가 실행될 때만 실제 값으로 바꿔 넣는 플러그인

Cross-project dependency, container-image and end-of-life tracking from Vigilant: an MCP server for project status and resolutions, a session-start brief when the current project has critical or high findings, and a skill for working through them. Self-gating: silent outside registered projects. Requires the vigilant command (uv tool install from samalone/vigilant).

Redacts secrets (Supabase service_role JWTs, API keys, private keys, .env and JSON config secrets) from tool results and conversation rows before the model reads them or the transcript stores them