Dechonet MCP

Phishing / Smishing Link Check

phishing_link_check

Read-onlyIdempotent

Check links in a suspicious text or email before tapping them: extract URLs, follow shorteners to real destinations, and judge phishing signals like fake bank domains or APK downloads.

Instructions

Check the links in a suspicious text message (smishing) or e-mail before anyone taps them. Paste the whole message or a single URL: every link is extracted, followed through URL shorteners to its real destination, and judged on observable signals — domain age from RDAP, impersonation of Korean banks/couriers/telecoms/portals (brand keyword, lookalike spelling, or the brand domain worn as a subdomain), Android APK downloads, punycode, public phishing/malware feeds, and DechoNet's own impersonation tracking. Use this when a user asks whether a link or message is a scam. Verdicts: danger, suspicious, unreachable, official (lands on the brand's own domain), no_signals (no warning signs found — never present this as "safe"). Read-only: page bodies are never run and the message text is not stored; requires no API key; rate-limited. Returns per-link verdicts with reasons and a shareable report link.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`text`	Yes	The suspicious message exactly as received (any language), or just the URL. Up to 5 links are checked.

Output Schema

TableJSON Schema

Name	Required	Description
`links`	Yes
`verdict`	Yes	danger \| suspicious \| unreachable \| official \| no_signals \| no_links
`reportUrl`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Addedv1.3.1Oct 6, 2026

TDQS

A4.6/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare read-only, idempotent, non-destructive, open-world, and the description still adds substantial behavior: shortener unwinding, RDAP domain-age lookups, APK/punycode checks, feed sources, no API key, rate limiting, and no storage of message text or execution of page bodies. It also defines the verdict vocabulary and warns that 'no_signals' must not be presented as safe.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the action and audience, then the mechanism, then the verdict list. It is dense but every clause carries information; the enumeration of signals is long, though it earns its place by explaining what the verdicts rest on.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists so return values need not be described, yet the description still covers the verdict taxonomy needed to interpret results, plus the critical 'never call no_signals safe' caveat, privacy, auth, and rate-limit facts. Nothing an agent needs to invoke or relay results is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the schema already explains the single 'text' parameter, up to 5 links. The description adds extraction semantics — paste the whole message or a single URL, and every link is extracted and followed — which slightly extends the schema's meaning.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb (check/extract/follow/judge) on a specific resource (links in a suspicious SMS or email), with concrete detection signals. It is clearly separable from siblings like lookalike_domains or impersonation_exposure, which look up brand domains rather than judging a pasted message.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives a clear triggering condition: 'Use this when a user asks whether a link or message is a scam,' plus the intent framing 'before anyone taps them.' It never names a sibling as the alternative or states when not to use it, so it stops short of the explicit routing seen in the top tier.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.