도구

Plugin marketplace from security firm Trail of Bits with 40+ plugins for code auditing, CodeQL/Semgrep static analysis, variant analysis, supply-chain risk, smart contract security and reverse engineering. Codex can load it directly too.

From Trail of Bits: reviews PRs, commits or diffs from a security angle, tracing history with git blame, estimating blast radius by caller count and checking test coverage gaps, then writes a Markdown review report.

Trail of Bits static analysis skills: CodeQL for interprocedural taint tracking and data-flow analysis, Semgrep for rule-based scanning, plus parsing and deduplicating SARIF results. Requires CodeQL or Semgrep installed locally.

Trail of Bits skill that audits GitHub Actions workflows wired to AI coding agents (Claude Code Action, Codex, Gemini CLI and others) for prompt injection, env-variable relays, dangerous sandbox settings, wildcard allowlists and other privilege risks.

From Trail of Bits: starts from one confirmed vulnerability, distills its root cause, then writes and progressively generalizes CodeQL/Semgrep queries to find similar variants across the codebase and report them by severity.

Trail of Bits skill that audits npm, PyPI and Go supply-chain risk: known advisories across direct deps and the whole lockfile tree, deprecated or archived upstreams, npm publisher concentration and install-time scripts. Needs uv; a logged-in gh is recommended.

Query, modify and convert JSON, YAML, TOML, XML, CSV, HCL, INI and more with one selector syntax, editing config files in place or converting between formats, handy for agents handling structured data in scripts. A single dependency-free binary that works much like jq or yq.

Query cloud resources and SaaS APIs live with SQL: plugins map AWS, Azure, GCP, GitHub, Kubernetes and other services to tables you can select from directly, no ETL needed. A single binary with Postgres built in.

Diffusion model inference in pure C/C++: one sd-cli command runs SD, FLUX, Qwen Image, Wan and more locally to create or edit images and short videos, with Metal acceleration on Mac. No cloud needed; bring your own model weights and use a release binary or build from source.

Reproduces issues on the Android emulator: installs builds, drives the UI with adb, dumps the UI tree, takes screenshots, captures logcat, and gathers performance evidence with Simpleperf and Perfetto. OpenAI's official Codex plugin; needs the Android SDK.

Creates, edits and validates reproducible Python and SQL Jupyter notebooks: organizes cells as reports or tutorials, checks conclusions against actual output and reruns top to bottom before handoff. Maintained by OpenAI in the Data Analytics plugin of openai/plugins.

Build with OpenAI best practices inside Codex: call the OpenAI API, build and deploy agents with the Agents SDK, create ChatGPT apps and prepare submission materials, and create API keys directly. Maintained by OpenAI; Codex only.

Security scans for a whole repo or for changes like PRs and commits, with threat modeling, vulnerability triage, fixes and re-verification; reports export to JSON, SARIF and more. OpenAI's official Codex plugin; requires Codex Security access.

Tests and debugs a locally running frontend with Codex's Browser plugin or Playwright: checks blank pages, framework and console errors, captures screenshots, re-verifies after fixes and writes a QA report. Maintained by OpenAI; formerly playwright-interactive in openai/skills.

Scrapes web pages into Markdown or structured data, with full-site crawling, sitemaps, web search and interaction with dynamic pages. Maintained by Firecrawl; the cloud needs an API key (self-hosting is possible), and the keyless hosted endpoint only offers scrape, search and parse.

Anthropic's reference implementation for finding and fixing vulnerabilities: Claude Code skills for threat modeling, static scanning, triage and patching, plus an autonomous recon-to-patch pipeline (C/C++ memory bugs by default; needs Docker and gVisor). No longer maintained.

Open-source LLM evaluation platform: one opencompass command picks the model and datasets and runs the evaluation across hundreds of benchmarks. Works with local Hugging Face models and API models, with vLLM and LMDeploy for faster inference.

Hands your current browser session to the agent through a Chrome extension: read console logs and network requests, take screenshots, and run Lighthouse performance, accessibility and SEO audits without signing in again.

Converts and two-way syncs Jupyter notebooks with plain Python scripts or Markdown, so agents can edit research and teaching notebooks like ordinary text and diffs stay readable. Needs a local Python environment; install with pip.

Cleans a LaTeX project before arXiv submission: removes comments, auxiliary files and unused .tex files and images, and can compress images, so internal notes are not published with the source. From Google Research; installable with Homebrew on macOS.

Python library and CLI for extracting main web content: fetches pages, crawls via sitemaps and RSS feeds, strips headers, footers and other noise, and outputs TXT, Markdown, JSON, XML and more with metadata. Also processes downloaded HTML offline.

Anthropic's official OS-level sandbox CLI: restricts file and network access for agents, MCP servers and any command without containers, using sandbox-exec on macOS and bubblewrap on Linux. Currently a research preview.

Static security analysis for CI/CD config: finds template injection, credential leaks, excessive permissions and more in GitHub Actions workflows, also covers Dependabot and pre-commit configs, and auto-fixes some findings.

Lets coding agents control real Chrome with short Puppeteer scripts, with named pages that keep state across calls, accessibility-tree snapshots and screenshots for self-testing web changes. Install the dev-browser CLI globally first; macOS and glibc Linux only.