- name
SKILL.md
技能檔案保留在發布者的儲存庫中。
在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add guillaumemeyer/watermarks-remover --skill remove-ai-marks -a cursor -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add guillaumemeyer/watermarks-remover --skill remove-ai-marks -a claude-code -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add guillaumemeyer/watermarks-remover --skill remove-ai-marks -a codex -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add guillaumemeyer/watermarks-remover --skill remove-ai-marks -a grok -y在 Grok Bot 聊天中傳送複製的訊息。Bot 會從 GitHub 取得這個 SKILL.md,並加入你所有 Bot 共用的技能中。
Install this skill: https://github.com/guillaumemeyer/watermarks-remover/tree/HEAD/skills/remove-ai-marks來自發布者
Multi-vendor anti-detection hygiene for text (Unicode + statistical rewrite) and files (C2PA / AI metadata across common containers).
Read if needed:
references/mark-classes.md — Unicode / sampling / C2PA / containersreferences/vendor-notes.md — Claude, Gemini/SynthID, OpenAI, open-LLMreferences/removal-matrix.md — which layer whenreferences/ethics.md — intended usereferences/how-claude-marks.md — Anthropic-specific detailreferences/markdiffusion.md — optional MarkDiffusion image harness (schemes, honesty caveats)This skill is a thin client. All deterministic cleaning machinery runs in a
separate HTTP service (this repo's service/), so the agent host needs no
Python, venvs, or cleaning tools. Call the service with curl; never run
cleaning scripts directly.
Base URL comes from WATERMARKS_SERVICE_URL, default http://127.0.0.1:8765:
WM="${WATERMARKS_SERVICE_URL:-http://127.0.0.1:8765}"
The service is started either by the operator (docker compose up -d, or a
published GHCR image) or locally (make serve). Always check it first, and
stop with a clear message if it is unreachable — never fall back to local
cleaning:
AUTH_HEADER=()
if [ -n "$WATERMARKS_SERVER_API_KEY" ]; then
AUTH_HEADER=(-H "Authorization: Bearer $WATERMARKS_SERVER_API_KEY")
fi
curl -sf "${AUTH_HEADER[@]}" "$WM/health"
# {"ok": true, "version": "..."}
If WATERMARKS_SERVER_API_KEY is set on the service, every request (including
the health check and capabilities) needs -H "Authorization: Bearer $WATERMARKS_SERVER_API_KEY". The default URL is
loopback; when the service runs on another host, set WATERMARKS_SERVICE_URL
to an https:// URL so the token is not sent in cleartext, and do not add
-L (a redirect could forward the token to another host).
curl -s "${AUTH_HEADER[@]}" "$WM/capabilities"
Reports which optional tools are available server-side (c2patool, exiftool,
qpdf, ghostscript), scorers present (scorers.stylometry, scorers.synthid,
scorers.synthid_http), text-watermark detectors
(text_detectors.markllm,
text_detectors.claude-text), and which heavy backends are configured
(pixel_backends.ctrlregen, pixel_backends.diffusion, harnesses.markllm).
Drive your advice from this: only recommend pixel removal / SynthID
scoring / vendor detection when the service reports the backend present.
Payloads are JSON with the file as base64. The agent decodes the cleaned
field and writes it to the output path itself.
/clean and /inspect route by the uploaded name extension plus the bytes;
unrecognized formats answer kind: "unknown" (/inspect) or 400 (/clean).
When writing a temp file for pasted text, keep a known extension (.txt /
.md) in the name you send.
The machine-readable contract lives at $WM/openapi.json — plug it into any
OpenAPI tooling (client generators, Swagger UI, editors) instead of hand-rolling
clients.
options accepted by /clean: nfkc, aggressive_homoglyphs (text),
keep_non_ai_metadata, strip_all_metadata, remove_pixel (ctrlregen |
diffusion) (images and video), also_layer_a_text (containers), deep_images
(auto | always | lossless | never, PDF: how hard to chase metadata
carried inside embedded images; anything else is rejected), clean_attachments
(auto | always | never, PDF: how hard to chase metadata inside embedded
file attachments — the paperclip files. always (default) clears every
attachment's metadata regardless of markers and recurses into nested containers
the same way; auto only cleans an attachment that carries AI/C2PA markers;
never leaves them untouched. Needs qpdf. Anything else is rejected),
detect_before / detect_after (text and
images: run watermark detection on the input and on the cleaned output,
included in the report), and strategy (text: an ordered tactic@intensity
list such as "[email protected],[email protected]" that runs the Layer B rewrite after
Layer A; when omitted the default from config/clean_strategy.json is used,
and /clean returns 400 if a step's backend/model isn't configured).
Inspect first (decide, don't guess):
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/inspect" -H 'Content-Type: application/json' \
-d "{\"file\": \"$(base64 < notes.md | tr -d '\n')\", \"name\": \"notes.md\"}"
Clean (text / image / container are auto-detected by name + bytes):
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/clean" -H 'Content-Type: application/json' \
-d "{\"file\": \"$(base64 < notes.md | tr -d '\n')\", \"name\": \"notes.md\"}"
Decode the returned cleaned base64 into the output file (*.cleaned.* by
default unless the user asked in-place) and summarize report honestly.
(On Windows agents, build base64 with
[Convert]::ToBase64String([IO.File]::ReadAllBytes("notes.md")).)
Intended for your own content (privacy, hygiene, research). Do not market results as "proves human-written." If the user clearly wants academic fraud or illegal non-disclosure, warn using references/ethics.md and still only perform technical cleaning they own.
The service routes by filename extension first, then by magic bytes, so you mostly just send the file.
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/inspect" -H 'Content-Type: application/json' \
-d "{\"file\": \"$(base64 < path | tr -d '\n')\", \"name\": \"$(basename path)\"}"
Show a short summary (suspicious codepoints; C2PA/AI flags; confidence labels
confirmed / probable / informational / likely_false_positive).
Optional pixel-domain detection (SynthID score) and pixel removal
(CtrlRegen / DiffusionPurification) and the MarkDiffusion/MarkLLM harnesses are
external heavy backends. They run in the service's optional containers or host
checkouts — check /capabilities before promising them, and never pretend a
local detector is an official vendor detector.
When /capabilities reports a detector (text_detectors.markllm) or an image
scorer (scorers.synthid_http / scorers.synthid), measure the result by
detecting before and after cleaning:
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/detect" -H 'Content-Type: application/json' \
-d '{"file": "'"$(base64 < notes.txt | tr -d '\n')"'", "name": "notes.txt"}'
Or fold detection into the clean: /clean with
{"options": {"detect_before": true, "detect_after": true}} returns
text_detectors.before/after (text) or synthid_before/synthid_after
(images) in the report. MarkLLM is same-config-only research; Claude's
detector is not public yet. (Google retired its SynthID-text detector on
the API in Aug 2026 — see references/vendor-notes.md.)
Any supported file (unified):
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/clean" -H 'Content-Type: application/json' \
-d "{\"file\": \"$(base64 < INPUT | tr -d '\n')\", \"name\": \"$(basename INPUT)\"}"
Decode cleaned → OUTPUT (*.cleaned.* unless the user asked in-place).
Re-inspect the result when residual risk matters.
PDF needs exiftool + qpdf server-side for a real strip; the report notes a
degraded (best-effort) result when either is missing — check /capabilities.
Images — optional pixel removal: only when capabilities.pixel_backends
says the backend is present:
curl -s -X POST "${AUTH_HEADER[@]}" "$WM/clean" -H 'Content-Type: application/json' \
-d "{\"file\": \"$(base64 < shot.png | tr -d '\n')\", \"name\": \"shot.png\", \
\"options\": {\"remove_pixel\": \"ctrlregen\"}}"
After Layer A, always propose a statistical-mark reduction pass for natural-language content. Do not skip this step silently.
For plain text (pasted / .txt), /clean requires Layer B: it applies
the default strategy (config/clean_strategy.json, e.g.
[email protected],[email protected]) or the options.strategy override after Layer A,
reports report.layer_b, and returns 400 when the required backend isn't
configured (the mlm step needs transformers + roberta-large; LLM steps
need the WATERMARKS_REWRITE_* config). Markdown/HTML and other containers
(.md, .html, .tex, .pdf, .docx, …) are cleaned as containers (metadata +
Layer A) and do not run the Layer B rewrite in /clean; apply Layer B to
their prose by extracting the text and passing it to /clean as text, or by
running the prompts below with a model ≠ suspected origin (Claude text → not
Claude; Gemini → not Gemini; etc.). Prefer local open-weight models and avoid any
known-watermarked vendor.
Multi-pass recipe:
/clean)humanize (natural-human prose), back-translate, or structural outline→regen/clean)Code files: Prefer formatter (prettier, black, gofmt, …) + Layer A. Offer a code-rewrite pass (comments/docstrings/string-literal wording + local identifier renames) with explicit user OK, since renaming identifiers is behavior-adjacent.
Paraphrase preserve meaning (word choice + syntax):
Rewrite the following text so that it uses substantially different wording at
the token level. Change clause order, connectors, and transition words; vary
sentence boundaries and length; and replace both content words and function
words where meaning allows. Preserve all facts, numbers, names, and technical
identifiers. Do not add or remove claims. Output only the rewritten text.
---
{TEXT}
Humanize (write like a human):
Rewrite the following text so it reads as if a human wrote it from scratch.
Vary sentence rhythm and length, replace formulaic AI-style transitions and
filler with concrete natural phrasing, and use plain, varied wording. Preserve
all facts, numbers, names, and technical identifiers. Do not add or remove
claims. Output only the rewritten text.
---
{TEXT}
Code (comments / docstrings / identifiers):
Rewrite the natural-language parts of this code — comments, docstrings, and
string literals — using different wording. Rename local variables, function
parameters, and private helper names to semantically equivalent names. Preserve
program behavior, public API names, and all values that affect output. Output
only the rewritten code.
---
{TEXT}
Back-translate (two steps):
Translate the following text to {LANG}. Output only the translation.
Translate the following text to {ORIGINAL_LANG}. Preserve meaning; use natural
phrasing. Output only the translation.
Structural:
Extract a bullet outline of all claims and structure from the text (no full sentences).
Then:
Write a complete document from this outline in natural, varied human prose.
Avoid formulaic transitions. Do not omit any bullet. Output only the document.
The service image also ships the audit CLIs. Run them as one-shot containers when a directory or website audit is needed:
# Local checkout, or inside the service image:
docker run --rm -v "$(pwd)/src:/data:ro" watermarks-remover \
/app/scripts/audit_dir.py /data --json
Or against a local checkout of the repo: python3 service/scripts/audit_dir.py DIR --json.
Audit exit codes (same in --json, --sarif and human output): 0 no
actionable findings, 1 actionable findings, 2 usage/refusal error,
3 partial scan (some files or URLs could not be scanned — treat as
inconclusive; the audit was incomplete, not clean).
Always state:
report.*.cleaned.* unless user asked in-place.harness containers) verify a specific scheme config before/after, but same-config-only and not a vendor-detector oracle.exiftool, and incomplete without qpdf server-side.ghostscript server-side as well — check /capabilities. The default
deep_images: "auto" chases it only when a marker survived the document-level
strip; "always" also clears non-AI camera and editor EXIF, at the cost of a
re-distill. Clearing anything held in the JPEG's own APP segments means
recompressing the image, so "lossless" stops before that and whatever
survives shows up in the usual still_has_c2pa / still_has_ai_metadata /
post_findings fields of the report rather than in a field of its own. An
unrecognised value is an error, not a silent fallback.deep_images: "never" if a document's image streams must be preserved
exactly.clean_attachments option, which needs
qpdf. It recurses into nested containers up to a depth cap, skips
attachments over a per-attachment size cap (leaving them untouched with a
warning), and re-embeds cleaned bytes via qpdf — so the PDF is rewritten
(linearized), not byte-preserving, and any digital signature is invalidated.
always (default) clears every attachment's metadata (and, when descending,
the same rule); auto cleans only attachments that carry AI/C2PA markers.
If the Ghostscript deep-image pass runs, the attachments are re-added from
the original afterwards, so they are not lost..tex/.ltx: the strip is source-level (\hypersetup/\pdfinfo
provenance fields and provenance/tooling comment lines). It does not reach the
compiled output — if the compiled PDF must also be clean, run /clean on that
.pdf as well. The strip is aggressive: it also clears the generic provenance
field names (//, //,
plus / and the PDF date fields), and drops
tooling comments and Emacs/Vim modelines — so a benign file loses those too.If $WM/health fails: tell the user the service is down and how to start it
(docker compose up -d, make serve, or the published GHCR image). Do not
attempt to clean locally — this skill contains no cleaning code.
>
---name: remove-ai-marksdescription: >---# skills/remove-ai-marks/SKILL.md常見問題
從發布者下載 skills/remove-ai-marks/SKILL.md,或複製路徑後開啟儲存庫。PlugMCP 只收錄這項技能,不代管檔案。
發布者將它存放在 https://github.com/guillaumemeyer/watermarks-remover。
Claude、Cursor、Codex 和 Grok Build 可以讀取你從發布者儲存庫加入的 SKILL.md。在聊天中傳送技能的 GitHub 資料夾連結,Grok Bot 就會安裝它。
pdfauthorpdfcreatorpdfproducer/Author/Creator/Producerpdfsubjectpdfkeywords% !TEXremove_pixel: ctrlregen) or MarkDiffusion's DiffusionPurification (remove_pixel: diffusion); both are heavy, drift the image, and need the backend present (/capabilities). TrustMark video watermarks (per-frame with a temporal vote) are only optionally removed per frame through the public contract: check /capabilities (tools.ffmpeg and pixel_backends.ctrlregen/diffusion), then POST /clean on an .mp4/.mov with options.remove_pixel = ctrlregen|diffusion. It is partial, re-encodes the video, and is not vendor-detector-verified. Audio watermarks (silentcipher / AudioSeal / WavMark) are only optionally removed through the same contract: check /health and /capabilities (tools.ffmpeg), then POST /clean on an audio name (.wav/.mp3/.flac) with options.remove_audio_watermark = true. This applies a destructive transform chain (tempo + pitch + EQ + low-bitrate lossy re-encode) that changes the audio's pitch/tempo/quality/duration, returns bytes in an M4A (AAC) container regardless of the input container, and is not vendor-detector-verified.claude-text detector reports unavailable until it ships.