righettod
UserCustomized toolbox containing materials to perform a secure code review with the help of a coding assistant.
Categories
Indexed Skills (10)
codebase-hotspotsv1
Generate a list of locations in the codebase where risky processing is performed from a security perspective.
codebase-hotspotsv2
Generate a list of locations in the codebase where risky processing is performed from a security perspective.
agent-csv-injection
Taint-analysis agent specialized in CSV injection (formula injection). Receives source code of functions along a data-flow path and determines whether user-controlled values written into CSV output are missing the mandatory single-quote prefix guard on the six formula-trigger characters. Returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
agent-generic
General-purpose taint-analysis agent. Receives source code of functions along a data-flow path and determines whether user-controlled input reaches any risky sink not covered by a dedicated agent. Applies the full Risky processing list and returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
agent-hash-input-ambiguity
Taint-analysis agent specialized in hash input ambiguity. Receives source code of functions along a data-flow path and determines whether multiple user-controlled values are concatenated without an unambiguous separator before being passed to a cryptographic hash function. Returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
agent-jwt-validation
Taint-analysis agent specialized in insecure JWT access token validation. Receives source code of functions along a data-flow path and determines whether JWT tokens are validated according to all mandatory security rules (algorithm confusion, missing claim checks, forbidden header acceptance, size limit, revocation). Returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
agent-log-forging
Taint-analysis agent specialized in log injection and log forging. Receives source code of functions along a data-flow path and determines whether user-controlled input reaches a logging call without neutralizing newline characters (enabling fake log entries) or without HTML-encoding (enabling XSS in web-based log viewers). Returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
agent-redos
Taint-analysis agent specialized in ReDoS (Regular Expression Denial of Service). Receives source code of functions along a data-flow path and determines whether user-controlled input reaches a regex engine in a way that enables catastrophic backtracking. Returns structured findings per .claude/skills/codebase-hotspotsv2/shared-rules.md.
codebase-overview
Generate a visual overview of a codebase as a Mermaid flowchart that maps each entry point directly to its final processing point (custom code or a third-party library). Use when the user wants to see, at a glance, where input enters a codebase and where it ultimately gets processed. Works for any common programming language. Optionally scope the analysis to a single entry point or package.
codebase-semgrep-findings-review
Review a Semgrep SARIF or JSON output file, validate each finding using semantic code analysis, and produce a filtered report with CONFIRMED, FALSE_POSITIVE, or NEEDS_HUMAN_REVIEW verdicts. Drops false positives that pattern-based rules cannot distinguish from real issues.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.