_index

A comparison guide for choosing AI coding agents to fix traditional software and repository vulnerabilities. It covers Codex, Claude Code, Cursor, Copilot, Devin, Shiba Studio, Hermes, and OpenClaw.

In plain words
What is it for?
It is for selecting and configuring a governed coding agent for vulnerability remediation, with bounded instructions and review steps.
Why use it?
It helps match an agent to where security findings come from, where code runs, and how fixes are reviewed.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/stevologic/security-recipes.ai/_index
Clone the repo
git clone --depth 1 https://github.com/stevologic/security-recipes.ai
Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,381 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00035 $0.04381
Opus 5 $0.00017 $0.02191
Sonnet 5 $0.00007 $0.00876
Haiku 4.5 $0.00003 $0.00438

Measured yesterday against content hash 1db1051f93a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

_index scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

CLI-only install uses `curl` / `iex`; after that, `hermes desktop` launches
content/agents/_index.md · 313 lines

How it starts

The opening of the file, as written. The whole thing — 313 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Use this page to choose a governed agent that matches your finding source, execution environment, and review path, then configure its native instruction surface. This is a workflow-fit comparison for AI agents remediating traditional software and repository vulnerabilities, not a model-quality benchmark or a guide to vulnerabilities in AI systems. If the system under review is the agent itself, including its tools, identities, memory, retrieved context, browser, MCP or A2A connections, and incident controls, use [AI Agent Security]({{< relref "/agentic-security" >}}). For integration architecture that delivers recipes, finding evidence, and policy to an existing agent, use [AI agent security-context integration]({{< relref "/docs/agent-integration" >}}). For the end-to-end method, return to [AI vulnerability remediation playbooks]({{< relref "/security-remediation" >}}).

Last updated August 21, 2026. Capabilities and documentation were verified against the linked primary sources. Rechecked August 23, 2026 against the same official pages. Stephen M Abbott maintains this workflow-fit comparison with Security Recipes contributors in the public source and revision history. See the review methodology and corrections policy.

Start with the agent your team already uses. The remediation pattern works across tools as long as the agent gets three things:

  • A local rule file or knowledge entry.
  • The specific security recipe for the finding.
  • Only the context needed to produce a PR or triage note.

{{< callout type="info" >}} Do not switch agents just for a recipe. The best first agent is usually the one already connected to your repos, approvals, and review habits. {{< /callout >}}

Supported agents

{{< cards >}} {{< card link="/github_copilot/" title="GitHub Copilot" subtitle="Use repository instructions, narrow issues, code-scanning alerts, and the GitHub Copilot cloud agent." >}} {{< card link="/claude/" title="Claude Code" subtitle="Use CLAUDE.md, rules, skills, hooks, and Claude Security for repository analysis and separately reviewed patches." >}} {{< card link="/cursor/" title="Cursor" subtitle="Use Cursor Rules, local agents, Cloud Agents, dependency automation, and Security Review where available." >}} {{< card link="/codex/" title="Codex" subtitle="Use AGENTS.md, skills, and bounded tasks across the app, IDE, terminal, cloud, CI, and Codex Security." >}} {{< card link="/devin/" title="Devin" subtitle="Use Knowledge, Playbooks, and repository .agents/skills/your-skill/SKILL.md files for hosted alert-or-ticket-to-PR workflows." >}} {{< card link="#shiba-studio" title="Shiba Studio" subtitle="Use per-agent instructions, integration scopes, skills, and MCP servers on local worktree-bound agents powered by Grok/xAI." >}} {{< card link="#hermes-desktop" title="Hermes Desktop" subtitle="Use skills and sandboxed execution backends (local, Docker, SSH, Daytona, Singularity, or Modal) with Nous Research's self-improving open-source agent." >}} {{< card link="#openclaw" title="OpenClaw" subtitle="Use workspace AGENTS.md operating rules and SOUL.md boundaries with a local Gateway and a bring-your-own model." >}} {{< /cards >}}

Read the full file on GitHub · 313 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 313 lines · 35 tokens per session scan A 1db1051f93a1

Subscribe to this mod's changes

_index is an agent published in the GitHub repository stevologic/security-recipes.ai (1 stars, last pushed yesterday), licensed Apache-2.0. It adds 35 tokens to every session and 4,381 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.