agentsclimarketplace

Defending applications

Skill telagod/code-abyss/skills/defending-applications

Application security defense knowledge for builders. Covers Web/API/GraphQL hardening (XSS/SQLi/SSRF/IDOR/BOLA/Mass Assignment/deserialization/upload/path traversal), authentication/authorization (OAuth 2.0/OIDC/JWT/Session/Cookie/SAML/SSO), and LLM application security (prompt injection, jailbreak, RAG poisoning, agent privilege escalation, output filtering). Use when designing or reviewing application-layer defenses, fixing CVE-class bugs in your own code, hardening auth flows, or threat-modeling LLM-powered features. For offensive testing see securing-systems/pentest, for incident response see securing-systems/blue-team, for infra-layer hardening see provisioning-infrastructure.From its SKILL.md

Install
npx -y skills add telagod/code-abyss --skill defending-applications

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

4.5 KB, ~1.4k tokens by cl100k_base, as published. Nobody here has run it

应用安全防御 · 镇魔甲

判断先于执行:决定「是否做 / 选什么 / 如何取舍」(栈、方案、架构、权衡)前,先读领域判断内核 skills/_kernel/security/SKILL.md——它管 judgment,本秘典管 execution;冲突时以内核判断为准。

开发侧防御秘典:站在 builder 一侧,把漏洞挡在合并前。 不教渗透,只教"如何写出杀不动的代码"和"如何把已生漏洞最小代价闭环"。 信级:项目源码/lock 文件 > 框架官方安全文档 > CVE/CWE > 训练记忆(标 [unverified])。

路由

意图加载触发词
Web/API/GraphQL 漏洞防御web-api-appsecXSS, SQLi, SSRF, BOLA, IDOR, Mass Assignment, GraphQL, 反序列化, 文件上传, 路径遍历, XXE, OWASP
认证授权与会话oauth-and-sessionsOAuth, OIDC, JWT, SSO, SAML, Session, Cookie, PKCE, refresh token, kid 注入
LLM 应用安全llm-appsecPrompt 注入, 越狱, 越权调用, RAG 投毒, agent tool 滥用, 输出过滤, jailbreak

何时使用

场景使用本 skill使用其他
review 自家代码、修 CVE、设计鉴权✅ 本 skill
写 SAST 规则、Semgrep 模式✅ 本 skill + securing-systems/code-audit
设计 LLM 应用的 guardrail✅ llm-appsec+ building-agent-systems/llm-security
红队渗透、写 PoC 攻击目标securing-systems/pentest
处理已发生的入侵、日志取证securing-systems/blue-team
容器/K8s/CI 加固provisioning-infrastructure
设计零信任、身份架构部分(OAuth/SSO 内)architecting-security

联动

  • securing-systems/code-audit:本 skill 给"怎么修",code-audit 给"怎么找"。
  • securing-systems/pentest:攻方视角验证本 skill 的修复是否真正闭环。
  • analyzing-security:自动化扫描胶水;本 skill 提供规则语义。
  • building-agent-systems/llm-security:agent 内部细节;本 skill 关注应用边界。
  • designing-architectures:威胁建模、API 设计;本 skill 关注落地代码。

执行链

威胁建模 → 入口梳理 → source→sink 追踪 → 防御层选型 → 最小修复 → 回归测试 → 检测信号埋点

每个漏洞的修复必须回答三问:

  1. 入口在哪:用户可控数据从哪进来(route/header/cookie/body/file/upstream API/RAG doc)。
  2. 信任边界在哪:何处由"untrusted"变"trusted",是否有显式校验。
  3. 检测信号是什么:修复后若被绕过,哪条日志/指标会响。

铁律

  1. 服务端为真:前端校验只是 UX,所有 authz/sanitize 必须在服务端再做一次。客户端代码即攻击面。
  2. deny by default:authz、CORS、CSP、SQL where 子句、文件类型——所有名单皆 allowlist 优先,blacklist 是反模式。
  3. 不信任 ID:URL/body 里的 object_id 永远先做 ownership 校验,再做业务逻辑。BOLA 是 API 漏洞之王。
  4. 可观测优先于完美防御:每个高风险路径必须有结构化日志 + 速率/异常告警,被绕过时能 5 分钟内发现。
  5. 修复要带回归:每个漏洞修复必须配 1 个攻击用例测试 + 1 个正常用例测试,否则六个月后必复发。

输出约束

  • 漏洞示例代码统一 ❌ 错代码 / ✅ 正代码 对比,错代码必须能跑通漏洞场景。
  • 攻击演示用 example.com / RFC 5737 网段(192.0.2.0/24);token/密钥用 <REDACTED>
  • 给修复方案时同时给:检测信号(log key / metric name)、回归测试骨架、性能/兼容回归点。

What ships with it: 3 files

23.7 KB alongside SKILL.md

Gives 0 of the 12 instructions most auth identity skills give in ~1.4k tokens

Counted across 408 of the 432 authors here whose files we hold, read 2026-09-06

  • Use parameterized queries for all SQL and database access to prevent SQL injectionin 60 of 408, across 40 files
  • Hash passwords with bcrypt or Argon2idin 50 of 408, across 34 files
  • Store secrets in Vault or environment variablesin 37 of 408, across 17 files
  • Add security headers to all responsesin 28 of 408, across 14 files
  • Rate limit public endpoints per clientin 22 of 408, across 8 files
  • Store tokens in httpOnly cookiesin 21 of 408, across 16 files
  • Use HTTPS in productionin 17 of 408, across 4 files
  • Apply rate limiting to all API endpointsin 17 of 408, across 9 files
  • Validate all input with Bean Validationin 16 of 408, across 3 files
  • Enable JWT or OIDC for stateless authenticationin 15 of 408, across 2 files
  • Scan dependencies for CVEsin 15 of 408, across 2 files
  • Log audit records for sensitive operationsin 15 of 408, across 2 files

Said here and by no other author read

  • Read the security judgment kernel before deciding
  • Verify object ownership before business logic
  • Apply minimal fixes
  • Pair each fix with regression tests
  • Add structured logging and anomaly alerts to high-risk paths
  • Show vulnerable and fixed code side by side

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.