Jeff Larson
Community@thejefflarson
Agent Skills by Jeff Larson
Showing 45 vetted skills indexed across 1 GitHub repositories.
thejefflarsonsecurity-misconfiguration
Identify insecure defaults and misconfigurations in server code and configurations.
thejefflarsonssrf
Detect and remediate SSRF vulnerabilities in outbound HTTP requests.
thejefflarsonprototype-pollution
Detect prototype pollution vulnerabilities in JavaScript and TypeScript code.
thejefflarsonthreat-model
Identify missing security controls and gaps in system design before implementation.
thejefflarsonhardcoded-secrets
Flag hardcoded API keys, passwords, tokens, and private keys in source code and configuration files.
thejefflarsonmcp-security
Validate MCP tool schemas and handlers against security risks.
thejefflarsonbroken-access-control
Enforce ownership gates and role-based middleware with 404 responses.
thejefflarsonsensitive-disclosure
Redact PII and credentials from inputs before LLM usage.
thejefflarsonmass-assignment
Filter request fields to prevent mass assignment vulnerabilities in web frameworks.
thejefflarsonauthentication-failures
Identify and remediate authentication failures in login flows, session management, and token handling.
thejefflarsoninsecure-plugin-design
Enforce input validation, per-tool authorization, and audit logging in LLM tool definitions.
thejefflarsonnosql-injection
Detect and mitigate NoSQL injection flaws in MongoDB and Elasticsearch queries.
thejefflarsonmulti-agent-trust
Enforce authentication, authorization, and scoped permissions on inter-agent communications.
thejefflarsonexcessive-agency
Gate high-impact agent actions behind human approval and audit logs.
thejefflarsonoauth-implementation
Enforce strict JWT validation and redirect URI matching in OAuth/OIDC implementations.
thejefflarsonhotspots
Identify security hotspots in a codebase and output a prioritized report.
thejefflarsongraphql-security
Identify and mitigate GraphQL security risks including unbounded query depth, introspection exposure, and batch attacks.
thejefflarsontraining-data-poisoning
Block poisoned or low-quality data from training and fine-tuning pipelines.
thejefflarsoninsecure-local-storage
Detect insecure local data storage patterns exposing credentials, tokens, and PII.
thejefflarsonpath-traversal
Detect and remediate directory traversal vulnerabilities across multi-language codebases.
thejefflarsoncsrf
Enforce CSRF tokens on state-changing endpoints across web frameworks.
thejefflarsonmodel-theft
Enforce authentication, rate limits, and response sanitization on inference endpoints.
thejefflarsonintegrity-failures
Detect unsafe deserialization patterns and unverified artifact trust in code and CI/CD workflows.
thejefflarsoninjection
Detect injection vulnerabilities in SQL, shell, and template contexts.