DiffiCat
REMOTE · APP.DIFFICAT.COM
Compare a website before and after a migration: missing pages, content changes, visual diffs.
Available components
Last scanned (17 hours ago)
Scores higher than 62% of remote servers
Reachable. Last reached 21 Nov 2026.
Category Testing and QA
Recent critical change
Authorization (30 Sept 2026). See the changelog before you install this server.
How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. How we score → Why this is hard to score →
Endpoint Security83
- The endpoint's TLS certificate is valid, in date, and uses a strong key. View diagnostics → Pass
- No authorisation is required to call this server. Nothing on it is marked destructive, but we can't show that an anonymous caller can only read: full credit needs the server to list at least one tool and every tool to be verifiably read-only. We use a best-effort check of tool names and descriptions to tell read-only tools from tools that change things, so some genuinely read-only tools may not get full credit. Names that start with a read verb (get, list, read, search and similar) and accurate readOnlyHint annotations help. Meet that to earn full credit, or require authorisation. See how to fix → View diagnostics → Partial
- HTTPS is enforced; there's no plaintext access path. View diagnostics → Pass
- The HSTS (Strict-Transport-Security) header is present. View diagnostics → Pass
- DNSSEC is configured correctly; the domain's records validate against the full chain to the root. View diagnostics → Pass
Transport & Reachability100
- Verified streamable-http transport via a live MCP handshake. View diagnostics → Pass
Schema Quality & AI Usability58
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 1837 tokens (~306/item across 6 items; 6 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management27
- Stability check failed: schema churn in the 30 days we've observed: 11 tool removals, 0 breaking changes, 0 auth/transport breaks, 2 additions. See how to fix → Fail
Tool Coverage100
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 100% of tool parameters carry a description.Pass
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.Pass
- An AI judge read all 7 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a current MCP spec version (2026-07-28).Pass
Checks we ran (13)
- TLS certificate
- Authorization
- HTTPS
- HSTS header
- DNSSEC
- Transport
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Does not apply to this hosted endpoint (2)
- Supply Chain SecurityScored on package channels only. This is a hosted endpoint, so there is nothing to install or inspect.
- Provenance & TransparencyScored on package channels only. This is a hosted endpoint, so there is nothing to install or inspect.
The score does not measure everything. What the score does not measure →
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 20 Nov 26 0
- Server version: 3.157.2 → 3.160.0 functional
- 15 Nov 26 0
- Server version: 3.157.0 → 3.157.2 functional
- 13 Nov 26 0
- The server rewrote its instructions, which are the text every model session reads security
- Tool “capture_screenshot” rewrote its description, which is the text the model reads security
- Tool “check_page” rewrote its description, which is the text the model reads security
- Tool “diff_deploy” rewrote its description, which is the text the model reads security
- Tool “get_balance” rewrote its description, which is the text the model reads security
- Server version: 3.154.0 → 3.157.0 functional
- 10 Nov 26 −1
No change was recorded against any check on this day. Stability & Change Management went from 33 to 23.
- 9 Nov 26 0
- Server version: 3.153.2 → 3.154.0 functional
- 8 Nov 26 0
- Tool “wait_for_comparison” rewrote its description, which is the text the model reads security
- Server version: 3.149.0 → 3.153.2 functional
- “capture_screenshot” reworded the description of “idempotencyKey” cosmetic
- “check_page” reworded the description of “idempotencyKey” cosmetic
- “check_page” reworded the description of “maxPriceUsd” cosmetic
- “diff_deploy” reworded the description of “idempotencyKey” cosmetic
- “diff_deploy” reworded the description of “maxPriceUsd” cosmetic
- “get_report” reworded the description of “failedOnly” cosmetic
- “get_report” reworded the description of “includeText” cosmetic
- “get_report” reworded the description of “jobId” cosmetic
- “get_report” reworded the description of “jobToken” cosmetic
- “get_report” reworded the description of “responseFormat” cosmetic
- “get_report” reworded the description of “view” cosmetic
- “wait_for_comparison” reworded the description of “jobId” cosmetic
- “wait_for_comparison” reworded the description of “jobToken” cosmetic
- 6 Nov 26 0
- We updated how we score, so this day's move reflects our rubric, not a change to the server See what changed → functional
- 4 Nov 26 −1
- The server rewrote its instructions, which are the text every model session reads security
- Tool “capture_screenshot” rewrote its description, which is the text the model reads security
- Tool “check_page” rewrote its description, which is the text the model reads security
- Tool “diff_deploy” rewrote its description, which is the text the model reads security
- Tool “get_balance” rewrote its description, which is the text the model reads security
- Server version: 3.148.0 → 3.149.0 functional
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 21 Nov 2026 · Probed https://app.difficat.com/mcp
TLS valid
Negotiated TLS 1.3 with TLS_AES_128_GCM_SHA256 .
| Subject | Issuer | Valid from | Valid until | Key | Signature | Serial |
|---|---|---|---|---|---|---|
| CN=app.difficat.com | CN=WR3,O=Google Trust Services,C=US | 23 Oct 2026 | 21 Jan 2027 | RSA 2048 | SHA256-RSA | 4a78711be2cf5ebf10070a7d7b1ab75d |
| SANs: app.difficat.com | ||||||
| CN=WR3,O=Google Trust Services,C=US (CA) | CN=GTS Root R1,O=Google Trust Services LLC,C=US | 13 Dec 2023 | 20 Feb 2029 | RSA 2048 | SHA256-RSA | 7ff005a91568d63abc22861684aa4b5a |
| CN=GTS Root R1,O=Google Trust Services LLC,C=US (CA) | CN=GlobalSign Root CA,OU=Root CA,O=GlobalSign nv-sa,C=BE | 19 Jun 2020 | 28 Jan 2028 | RSA 4096 | SHA256-RSA | 77bd0d6cdb36f91aea210fc4f058d30d |
Background: What to check on a remote MCP endpoint →
DNSSEC secure
Validation of app.difficat.com. — Secure
| Zone | DS | Keys | Algorithms | Outcome |
|---|---|---|---|---|
| . | Trust anchor | 20326, 38696 | 8, 8 | Verified |
| com. | DS published | 19718 | 13 | Verified |
| difficat.com. | DS published | 33021 | 13 | Verified |
| app.difficat.com. | Verified address RRset verified with the apex keys |
Authentication No authorisation required
The endpoint answered without asking for a token. Anyone who knows the URL can reach it.
| Result | No authorisation required |
|---|---|
| HTTP status | 200 |
| Header | Value |
|---|---|
| strict-transport-security | max-age=31536000; includeSubDomains |
| content-security-policy | default-src 'self';script-src 'self' https://eu-assets.i.posthog.com https://internal-cf.posthog.com;style-src 'self' 'unsafe-inline' https://eu-assets.i.posthog.com;img-src 'self' data: blob:;connect-src 'self' https://eu.i.posthog.com https://eu.posthog.com https://internal-cf.posthog.com;font-src 'self';object-src 'none';frame-ancestors 'none';base-uri 'self';form-action 'self';script-src-attr 'none';upgrade-insecure-requests |
| x-content-type-options | nosniff |
| x-frame-options | SAMEORIGIN |
| referrer-policy | no-referrer |
Background: How OAuth 2.1 works in the 2026 MCP spec →
Transports 2 probes
| Transport | URL | Outcome | Status | Location |
|---|---|---|---|---|
| streamable-http | https://app.difficat.com/mcp | Verified | 200 | |
| http (plaintext) | http://app.difficat.com/mcp | HTTPS enforced | 302 | https://app.difficat.com/mcp |
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
capture_screenshot Capture screenshot ~249
Takes a full-page screenshot of one URL, for example to see how a relaunched page or a mobile layout looks. Free 10 times per client address per day, no key needed. The PNG is taken at the given viewport width and returned inline when small, else as links.file to fetch with the jobToken. Call it for a one-off look at a page or a mobile-width check. Do not call it for pages a comparison job covers: the job already stored their screenshots. Example: { "url": "https://new.example/pricing", "viewportWidth": 375 }. Answers with jobId, jobToken, filename, url, dimensions, bytes, links.file and, on a free run, sample and allowanceRemaining.
| Name | Type | Req | Description |
|---|---|---|---|
| auth | object | – | HTTP basic auth for the page. |
| idempotencyKey | string | – | Retry key: under a prepaid balance, the same value within 24 hours is not charged again and answers DUPLICATE_REQUEST with the first jobId. |
| url | string | yes | Page to capture (full URL). |
| viewportWidth | integer | – | Capture width in px. |
| visual | object | – | Capture rules applied to both sides. |
| Name | Type | Req | Description |
|---|---|---|---|
| allowanceRemaining | integer | – | – |
| bytes | integer | – | – |
| code | string | – | – |
| dimensions | object | – | – |
| error | string | – | – |
| filename | string | – | – |
| followUps | array | – | – |
| imageSize | object | – | – |
| jobId | string | – | – |
| jobToken | string | – | – |
| links | object | – | – |
| pagesCharged | integer | – | – |
| payment | – | – | – |
| region | object | – | – |
| sample | boolean | – | – |
| url | string | – | – |
No examples provided.
check_page Check page ~255
Use when you deployed or edited a page and need to catch visual regressions and content changes: it compares the live URL with a baseline captured on the first call, a snapshot or a production URL, and answers the regressions. Free up to 100 pages per client address per day, no key needed.
| Name | Type | Req | Description |
|---|---|---|---|
| baseline | – | – | What to compare against: a saved snapshot ({ snapshotId, shareToken? }) or a production origin ({ productionUrl }). Omit it to use the baseline the first call captured. |
| checkMode | string | – | full (existence, content and visual, the default), text-only (no screenshots) or existence-only. |
| idempotencyKey | string | – | Retry key: under a prepaid balance, the same value within 24 hours is not charged again and answers DUPLICATE_REQUEST with the first jobId. |
| maxPriceUsd | number | – | Spend cap in USD: a job priced above it is refused before any challenge. |
| rebaseline | boolean | – | Capture a new baseline now instead of comparing with the stored one. |
| url | string | yes | The live page to check (full URL). |
| viewportWidth | integer | – | Capture width in px (default 1440). |
| Name | Type | Req | Description |
|---|---|---|---|
| allowanceRemaining | integer | – | – |
| baseline | object | – | – |
| code | string | – | – |
| error | string | – | – |
| followUps | array | – | – |
| httpStatus | integer | – | – |
| jobId | string | – | – |
| jobToken | string | – | – |
| ok | boolean | – | – |
| pagesCharged | integer | – | – |
| pagesChecked | integer | – | – |
| payment | – | – | – |
| regressions | array | – | – |
| sample | boolean | – | – |
| status | string | – | – |
| url | string | – | – |
No examples provided.
diff_deploy Diff deploy ~288
Use when a staging or preview deploy should match production, or a migrated site its old one: it compares the pages of both origins and answers only the pages that are missing, changed or look different. Free up to 100 pages per client address per day, no key needed.
| Name | Type | Req | Description |
|---|---|---|---|
| checkMode | string | – | full (existence, content and visual, the default), text-only (no screenshots) or existence-only. |
| idempotencyKey | string | – | Retry key: under a prepaid balance, the same value within 24 hours is not charged again and answers DUPLICATE_REQUEST with the first jobId. |
| maxPages | integer | – | Without paths: pages discovery compares, default 10. |
| maxPriceUsd | number | – | Spend cap in USD: a job priced above it is refused before any challenge. |
| paths | array | – | Pages to compare, each starting with /; skips discovery. |
| previewUrl | string | yes | The preview or staging site (full URL), compared as Site B. |
| productionUrl | string | yes | The production site (full URL), compared as Site A. |
| threshold | number | – | Visual diff tolerance as a fraction from 0 (exact) to 1: 0.05 means 5%. The web app takes a percentage; the API does not. |
| viewportWidth | integer | – | Capture width in px. |
| Name | Type | Req | Description |
|---|---|---|---|
| allowanceRemaining | integer | – | – |
| code | string | – | – |
| error | string | – | – |
| followUps | array | – | – |
| jobId | string | – | – |
| jobToken | string | – | – |
| ok | boolean | – | – |
| pagesCharged | integer | – | – |
| pagesChecked | integer | – | – |
| payment | – | – | – |
| regressionCount | integer | – | – |
| regressions | array | – | – |
| sample | boolean | – | – |
| status | string | – | – |
No examples provided.
get_balance Get prepaid balance ~144
The prepaid balance and last 20 ledger entries of the key in the request’s Authorization: Bearer or X-DiffiCat-Key header; a human obtains one on the account page (/account), which a refusal’s nextSteps name. A Pro key also returns allowance (included pages shared with the web app, used, reset time) and overage (amount this period, cap). Each entry names its surface, web or api. Without a key, answers the free allowance left for your address and prices, the USD a call costs beyond it; comparisons, discoveries and screenshots on old.difficat.com and new.difficat.com draw on demoPages first. Free.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| allowance | object | – | – |
| balance | string | – | – |
| code | string | – | – |
| currency | string | – | – |
| error | string | – | – |
| free | object | – | Per kind: remaining, total and resetsAt, the time the kind renews |
| keyId | string | – | – |
| nextSteps | object | – | – |
| overage | object | – | – |
| prices | object | – | USD per compared page (perPageTier2Usd past tierPages), discovery and screenshot |
| recent | array | – | – |
No examples provided.
get_report Get report ~420
The finished report. Call it first with failedOnly: true and view: "digest": only the failed pages, each with its issues, diff percentage and largest regions, and a text line per page; summary covers every page. check_page and diff_deploy return the first regressed region's image; get_screenshot (?surface=full) fetches a region by filename. Without view, results are { path, status, failedChecks }; responseFormat: "detailed" returns per-check fields. offset and limit page through results (default 50, at most 5000; nextOffset is next); limit: 1 returns the summary alone; includeText: true adds the plain-text report. A change whose flakyShare is at least 0.5 is likely the page moving on its own: mask its fingerprint's element (visual.masks), not a regression. Call wait_for_comparison first. Free.
| Name | Type | Req | Description |
|---|---|---|---|
| failedOnly | boolean | – | Only failed pages in results, total, nextOffset and text; summary still covers every page. |
| includeText | boolean | – | Also put the text block in structuredContent.text. |
| jobId | string | yes | Returned at job creation, e.g. web-1a2b3c. |
| jobToken | string | – | Returned at job creation; optional when the HTTP headers carry the key that paid the job. |
| limit | integer | – | Page results to return (at most 5000; default 1000 on REST, 50 over MCP). |
| offset | integer | – | Index of the first page result to return (default 0). |
| responseFormat | string | – | concise (default): each result is { path, status, failedChecks }, plus snapshots when a side came from one; detailed: every per-check field, diff and screenshot name. |
| view | string | – | full (default): every page result with its diffs, screenshots and text; digest: one small object per page (status, failed checks, issues, diff percentage), no text. |
| Name | Type | Req | Description |
|---|---|---|---|
| cancelled | boolean | – | – |
| code | string | – | – |
| error | string|null | – | – |
| followUps | array | – | – |
| jobId | string | – | – |
| nextOffset | integer | – | – |
| results | array | – | – |
| status | string | – | – |
| summary | object | – | – |
| text | string | – | – |
| total | integer | – | – |
No examples provided.
wait_for_comparison Wait for comparison ~137
Waits until the job is completed, failed or cancelled, or timeoutSeconds pass, then answers status, progress and, once terminal, the summary (passed, failed, issuesBySeverity). On "running", call it again. Free.
| Name | Type | Req | Description |
|---|---|---|---|
| jobId | string | yes | Returned at job creation, e.g. web-1a2b3c. |
| jobToken | string | – | Returned at job creation; optional when the HTTP headers carry the key that paid the job. |
| timeoutSeconds | integer | – | Seconds to wait for a terminal status before answering with the current one (default 60, at most 300). |
| Name | Type | Req | Description |
|---|---|---|---|
| code | string | – | – |
| error | string | – | – |
| followUps | array | – | – |
| jobId | string | – | – |
| pagesReturned | integer | – | – |
| progress | object | – | – |
| status | string | – | – |
| summary | object | – | – |
No examples provided.
How do I install the DiffiCat MCP server?
DiffiCat is a hosted endpoint at https://app.difficat.com/mcp, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
remote · app.difficat.com
claude mcp add --transport http com-difficat-difficat 'https://app.difficat.com/mcp'
What is the DiffiCat MCP server?
DiffiCat is an MCP server listed in the public MCP registry as com.difficat/difficat. Compare a website before and after a migration: missing pages, content changes, visual diffs. This page covers its hosted endpoint (https://app.difficat.com/mcp).
Is the DiffiCat MCP server safe to use?
DiffiCat scores 75 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the DiffiCat MCP server expose?
DiffiCat exposes 6 tools: check_page, diff_deploy, wait_for_comparison, get_report, capture_screenshot, get_balance. Their descriptions and schemas cost roughly 1,493 tokens of context every time the server is loaded.
Does the DiffiCat MCP server require authentication?
No. We connected to DiffiCat without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.
Is the DiffiCat MCP server still maintained?
DiffiCat is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What did VerifyMCP test on the DiffiCat MCP server?
VerifyMCP ran 13 checks on DiffiCat, covering TLS certificate, Authorization, HTTPS, HSTS header and 9 more. Supply Chain Security and Provenance & Transparency do not apply to a hosted endpoint, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.