BC hunt-xss
Hunting skill for xss vulnerabilities. Built from 174 public bug bounty reports. Use when hunting xss on any target. For markup injection that reflects raw HTML but does NOT execute JavaScript (no `<script>`/event-handler execution), see hunt-html-injection — escalate here once script execution is possible.
Hunting skill for xss vulnerabilities.
As a process C 50/100 · Has gaps — weak spots: result and completion, when it triggers, inputs and preconditions
ProcedureSoftware developmentData and analyticsSecuritytype and topics are labelled automatically from the skill text
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
For the model run — optional
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 7
✓ No critical or high findings
Medium and low: 7
-
low Risky intent
intent-offensive-securitySKILL.md:3Offensive-security / dual-use content (legitimate for authorised testing; review intended use)description: Hunting skill for xss vulnerabilities. Built from 174 public bug bounty reports. Use when hunting xss on any target. For markup injection that reflects raw HTML but does NOT execute JavaS
-
low Risky intent
intent-offensive-securitySKILL.md:50Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **Payment/financial flows** (`paypal.com`, checkout pages, currency converters) — XSS here enables credential harvesting and financial fraud at scale
-
low Risky intent
intent-offensive-securitySKILL.md:151Offensive-security / dual-use content (legitimate for authorised testing; review intended use)2. **Classify by type** — Determine if each reflection is Reflected (URL param → response), Stored (database → later rendering), or DOM-based (JS reads URL/storage → DOM sink). Each requires different
-
low Dangerous commands
cmd-eval-dynamicSKILL.md:322Dynamic code execution from decoded/untrusted input (documentation of a security skill)<svg onload=eval(atob('YWxlcnQoMSk='))>security skill -
low Risky intent
intent-offensive-securitySKILL.md:369Offensive-security / dual-use content (legitimate for authorised testing; review intended use) (quoted — discussed, not commanded)An attacker discovered that a major payment platform's sign-in page reflected user-controlled input and was cached by the CDN layer. By sending a crafted request that poisoned the cache, the attacker
quoted -
low Risky intent
intent-offensive-securitySKILL.md:428Offensive-security / dual-use content (legitimate for authorised testing; review intended use)- **Impact:** Privilege escalation from low-priv user to admin via stored XSS — attacker promotes themselves on the privileged user's behalf.
A further 1 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 1. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 7257 tokens (recommended < 5000); move details to references/ - note
frontmatter-keyunknown frontmatter key "sources" - note
frontmatter-keyunknown frontmatter key "report_count" - note
edit-residuethe text marks something as outdated (lines 303, 341): check that old rules are not kept next to new ones — the full check reads the text for contradictions
Process rating: all ten parameters 50/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 8 mutating operations with no state check
- 60Tools and files. Uses tools (bash, web) that frontmatter does not declare
- 60Failures and branches. 2 branches
- 70Execution cost. Instruction body is 7257 tokens
- 100Steps. 113 steps
- 100Consistency. Name and required fields are in place
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 12 top-level sections: this looks like several domains in one skill
- low The response is described with custom markup (33 tags): a typed call is more reliable
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +3Description length 308: enough signal without eating the budget
- +4Structure: 22 headings
- +3Step-by-step instructions: 113 items
- +4Has examples (23 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 72.