Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add UnboundCompute/security-agent-skills --skill hunting-reflected-and-stored-xssgit clone --depth 1 https://github.com/UnboundCompute/security-agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/unboundcompute/security-agent-skills/hunting-reflected-and-stored-xss)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/hunting-reflected-and-stored-xss"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/hunting-reflected-and-stored-xss/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/hunting-reflected-and-stored-xss"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/hunting-reflected-and-stored-xss.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00170 | $0.02133 |
| Opus 5.5 | $0.00068 | $0.00853 |
| Sonnet 5.5 | $0.00034 | $0.00427 |
| Haiku 4.5 | $0.00017 | $0.00213 |
Grade A, and why
hunting-reflected-and-stored-xss scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hunting reflected and stored XSS: when data becomes markup in the wrong context
A cross-site scripting bug is an encoding failure with a location. The server takes untrusted data, reflected straight from the request or read back from storage, and writes it into an HTML response. If the value is encoded correctly for the exact place it lands, it stays inert data; if it is not, it becomes markup or script and runs in the victim's browser with the victim's session. The subtlety is that "correctly encoded" depends entirely on the output context: the escaping that neutralizes a value in an HTML body does nothing for a value inside an unquoted attribute, an inline script, a URL, or a stylesheet. So the hunt is not "is this escaped" but "is this escaped for the context it is actually in." You find these by locating every place server-rendered output includes untrusted data and matching the encoding to the context, then following whether delivery is reflected from the request or stored and served later.
When to use
- A server-side template or string-built response embeds request parameters or stored data into HTML.
- The framework autoescapes but code uses a raw, safe, or unescaped marker in places.
- User-controlled data is serialized into JSON or an attribute that a page then interprets.
Scope check
Test XSS only against applications you own or are authorized to assess, with test accounts, and use a harmless proof such as a benign marker rather than a real payload against other users, because stored XSS executes in every viewer's session. Clean up any stored test value you inject. If you can't name the authorization, stop.
The loop
- Establish the exact output context and its encoding first. For each place untrusted data reaches the response, determine the precise context, HTML body, quoted attribute, unquoted attribute, inline script, URL, or CSS, and whether the encoding applied matches that context. This is the false-positive killer: HTML-entity encoding is correct in a body but useless inside a script block or an unquoted attribute, so an escaped value can still be XSS in the wrong place, and a value that never leaves a body with correct entity encoding is inert. Name the context and its encoding before judging.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 138 lines · 170 tokens per session scan A a47b58eb7162
hunting-reflected-and-stored-xss is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 1mo ago), licensed MIT. It adds 170 tokens to every session and 2,133 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-05.
Other skills, from other repositories
prodcheck-review
Review this codebase against the prodcheck pre-production checklists — security, performance, scale, integrations, post-launch readiness and people. Use when asked to check whether a project is ready to ship, to audit an area before launch, or to work through a specific checklist. Produces evidence with file:line…
integrate-arcjet-guard-genkit
Integrate Arcjet security into a Genkit JS agent using @arcjet/guard — wrap ai.defineTool, put guardMiddleware on generate({ use }) for unwrapped / MCP / filesystem tools, and read a caller-owned id from generate({ context }). Use when asked to add Arcjet to genkit, rate limit its tools, screen inbound messages, or…
integrate-arcjet-guard-tanstack-ai
Integrate Arcjet security into a TanStack AI chat() app using @arcjet/guard — put guardMiddleware first on chat({ middleware }) so onBeforeToolCall gates tools, and read a caller-owned id from chat({ context }). Use when asked to add Arcjet to TanStack AI, rate limit its tools, screen inbound messages, or block prompt…
integrate-arcjet-guard-claude-managed-agents
Integrate Arcjet security into Claude Managed Agents (hosted REST+SSE, beta managed-agents-2026-04-01) using @arcjet/guard — screen user.message / initialevents before sessions.events.send, and gate custom tools on agent.customtooluse. Use when asked to add Arcjet to Claude Managed Agents, rate limit custom tools, or…
linear-claude-skill
Manage Linear issues, projects, and teams.
varlock-claude-skill
Secure environment variable management ensuring secrets are never exposed in Claude sessions, terminals, logs, or git commits.