How do I install the Hunt Sqli skill?

Run npx skillmds@latest add duriantaco/hunt-sqli in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Hunt Sqli safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Hunt Sqli?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Hunt Sqli free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Hunt Sqli?

duriantaco (@duriantaco) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Hunt Sqli by duriantaco — Agent Skill

Name: Hunt Sqli
Author: duriantaco

Hunt Sqli

Investigate authorized SQL injection boundaries on web and API inputs. Use when typed discovery or target-origin evidence identifies query-backed parameters, database-error signals, boolean or timing SQL differentials, SQL-filter behavior, or a suspected injectable login, search, sort, filter, or identifier input.

duriantaco Updated Sep 17, 2026

File contents

Hunt SQL Injection Boundaries

Treat this card as advisory prioritization. Scope, replay contracts, TrafficPolicy, and native evidence gates remain authoritative.

Workflow

Select a query-backed input and preserve its method, identity, session, fixed fields, and encoding.
Run sqli_differential first; change one input and pair each candidate with a same-shape control.
Interleave bounded timing pairs. Use sqli_exploit only after a reproducible differential and filtered_query_bypass only when target evidence shows filtering.
Preserve redacted replays, deltas, controls, and evidence references. Report validated vulnerabilities independently of any challenge objective.

Evidence Gate

Confirm only a new database error absent from control, a reproducible boolean differential, or repeatable timing above baseline jitter. Vendor fingerprints, isolated errors, length changes, and one slow response remain observations. Keep contract_missing results suspected.

Stop Conditions

Stop without a replayable query input, stable control, or policy budget, or after minimal proof. Do not request extraction beyond the authorized native closer; redact credentials and bodies.

Frequently asked questions