FaintMark
PYPI · FAINTMARK
Invisible per-recipient image watermarks. Mark one copy per recipient, then identify a leaked copy.
Available components
Last scanned (16 hours ago)
Scores higher than 67% of PyPI servers
Category Security
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security99
- No malware found by supply-chain analysis.Pass
- Known CVEs were checked across the 42 of 43 dependencies we could resolve, so this covers what we could see, not the whole tree.Partial
- Runs hatchling.build at install time, a recognised build step with no custom scripting around it. View diagnostics → Pass
- Dependency health was assessed across the 42 of 43 dependencies we could resolve, so this covers what we could see, not the whole tree. View diagnostics → Partial
Provenance & Transparency48
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 5 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability61
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 763 tokens (~254/item across 3 items; 3 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 Management0
- Stability not yet verified: not enough scan history yet (needs a 30-day window).Unverified
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
- We read all 3 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 4 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Unverified: 1 category
A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.
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.
- 17 Oct 26 +15
- Malware scan: unverified → pass ▲ security
- 14 Oct 26 +14
- Malware scan: pass → unverified ▼ security
- Injection markers: unverified → pass ▲ security
- First check of Judged manipulation: pass security
- Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window). security
- Tool coverage: unverified → 100 ▲ functional
- MCP protocol: unverified → pass ▲ functional
- First check of Schema quality: fail functional
- First check of Destructive annotations: pass functional
- First check of Schema quality: fail functional
- First check of Schema quality: good functional
- First check of Tool coverage: 100 functional
- First check of Tool coverage: 100 functional
- 13 Oct 26 +10
- Known CVEs: unverified → partial ▲ security
- Dependency health: unverified → partial ▲ functional
- 12 Oct 26 29
First indexed and scored.
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 17 Oct 2026 · Analysed pypi/faintmark@0.1.0
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | pypi |
| Reason | No attestation published |
Background: How many MCP packages publish verified provenance →
Install scripts 1 script
| Hook | Tier | Command |
|---|---|---|
| build_backend | allowlisted | hatchling.build |
Background: Why install scripts are a supply-chain risk →
Dependencies 42 packages
| Packages resolved | 42 |
|---|---|
| Stale | 4 |
| Tree resolution | Partial |
The dependency tree was only partially resolved, so these counts may be incomplete.
Background: SBOMs and build attestations, explained →
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 →
check_image ~324
Decode the id in an image (any format Pillow opens, up to 100 megapixels) and look it up in the local registry; designed to work after chat-app recompression or resizing. status is matched (the id is registered to the returned recipient), unregistered (an id was read but it is not in this registry) or not_detected (no readable mark, which says nothing about where the image came from). A match also reports same_image when a stored hash exists, a perceptual-hash comparison with the issued copy: false means they differ beyond the threshold, which cropping or an id copied into another picture can cause. same_image true is not authentication: a tuned forgery passed it in our measurements. deep=true also tries a flip and small rotations (transform names the one that recovered the result), and when nothing is detected lists the registered ids within 26 of 96 codeword bits of the raw decoder output as suspects (up to three, each with id, recipient and bit_errors, plus same_image when a stored hash exists; empty when none is that close) and registry_size. Suspects are uncalibrated candidates by codeword distance: they can be coincidental (more likely with a large registry) or forged, and they never mean a match. files=true adds the original file name and the issued copy's path.
| Name | Type | Req | Description |
|---|---|---|---|
| deep | boolean | – | try flips and rotations, then list suspects |
| files | boolean | – | include file names and paths |
| image_path | string | yes | absolute path of the image to check |
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | – | – |
| marked_at | string | – | – |
| original | string | – | – |
| path | string | – | – |
| recipient | string | – | – |
| registry_size | integer | – | – |
| same_image | boolean | – | – |
| status | string | yes | – |
| suspects | array | – | – |
| transform | string | – | – |
No examples provided.
list_copies ~46
List registered copies, newest first. files=true adds file names and paths.
| Name | Type | Req | Description |
|---|---|---|---|
| files | boolean | – | include file names and paths |
| limit | integer | – | newest N copies |
| Name | Type | Req | Description |
|---|---|---|---|
| copies | array | yes | – |
No examples provided.
mark_image ~274
Create one invisibly watermarked copy of an image per recipient label and register them locally. Use before sharing a picture with several people so a copy that turns up later can be traced to the label it was issued under. Writes files only and sends nothing; keeps the source format (JPEG stays JPEG; HEIC becomes JPEG). Accepts JPEG, PNG, WebP or HEIC (HEIC only when the server runs with the heic extra) up to 25 megapixels, with no animation, transparency, CMYK or 16-bit grayscale, and 1 to 100 labels of 1 to 200 printable characters. Returns the output folder and each copy's path, id and warnings. A warning starting with 'may not survive' means the copy is fragile, so suggest sending it as a file rather than a photo; one starting with 'JPEG recompression not tested' means that check was skipped. Cancelling the call stops it at the next copy and removes the copies written so far; a batch already being registered completes.
| Name | Type | Req | Description |
|---|---|---|---|
| format | – | – | default: the source format |
| image_path | string | yes | absolute path of the image to mark |
| output_dir | – | – | default: a marked/ folder beside the image |
| recipients | array | yes | 1 to 100 labels, one copy each |
| Name | Type | Req | Description |
|---|---|---|---|
| copies | array | yes | – |
| original | string | yes | – |
| output_dir | string | yes | – |
No examples provided.
How do I install the FaintMark MCP server?
FaintMark runs locally as a PyPI package, launched with uvx faintmark. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
pypi · faintmark
claude mcp add matchalatteguy-faintmark -- uvx faintmark
What is the FaintMark MCP server?
FaintMark is an MCP server listed in the public MCP registry as io.github.matchalatteguy/faintmark. Invisible per-recipient image watermarks. Mark one copy per recipient, then identify a leaked copy. This page covers its PyPI package (faintmark).
Is the FaintMark MCP server safe to use?
FaintMark scores 68 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 FaintMark MCP server expose?
FaintMark exposes 3 tools: mark_image, check_image, list_copies. Their descriptions and schemas cost roughly 644 tokens of context every time the server is loaded.
Is the FaintMark MCP server still maintained?
FaintMark is still listed as active in the MCP registry. We last reached this channel on 17 October 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the FaintMark MCP server under?
FaintMark declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.