Grammarly Data Safety Guardian
Run a deterministic preflight over a metadata-only submission manifest. The guard
answers whether a planned document upload satisfies this skill's hard safety gates;
it does not read a document, contact Grammarly, create an upload URL, or retain a
copy of the manifest.
Overview
The guard is a closed-world policy check for a planned document submission. It
combines Grammarly's documented format and size constraints with local governance
decisions while keeping the document itself outside the workflow.
When to use
Use for a Writing Score, AI Detection (Beta), or Plagiarism Detection (Beta)
document submission before and after the evaluator inspects the provider-issued upload
origin. The preliminary decision gates OAuth metadata disclosure; the final decision
gates document bytes. Do not trigger for writing or editing content, document previews,
credential troubleshooting, or a request to upload a real document.
Prerequisites
- A synthetic or production-approved metadata-only manifest matching
references/submission-contract.md.
- Independent data-owner approval, consent evidence, transfer approval, content
classification, and acknowledgment of provider retention boundaries.
- Python 3.10 or newer. No document path, document content, credential, network, or
writable output directory is required.
Instructions
Collect only the manifest described in
references/submission-contract.md. The
manifest may contain extension, byte count, optional derived text counts, and
governance decisions, the pinned API control-plane origin, and the exact sanitized
presigned upload origin; it must not contain text, content, previews, credentials,
paths, filenames, full upload URLs, or raw API responses.
Run the guard with stdin:
python3 scripts/audit_submission_manifest.py < submission-manifest.json
Treat BLOCKED as a hard stop. Fix the named metadata or governance decision;
never truncate or transform a document silently to pass the guard.
Treat INSPECTION_READY as permission only to create a normalized-filename job and
inspect its sanitized upload origin. Treat READY as permission to continue to the separately authorized upload
workflow, not as proof that a document is lawful, accurate, or safe for a new
destination.
The script accepts strict JSON, rejects duplicate keys and non-standard constants,
recursively rejects raw text, location, response, and credential fields, and performs
no network or filesystem writes. It enforces the documented extensions, 4,194,304-byte
maximum, UTF-8 text counts, a non-restricted classification, all approval gates, the
exact https://api.grammarly.com control-plane origin, and an explicitly approved
public HTTPS S3 presigned-upload origin on the default TLS port.
Output
Return the deterministic JSON decision and failed check names. Do not echo a
filename, path, text, preview, credential, identifier, or raw document metadata
beyond the safe numeric and enum fields in the script's output.
Error Handling
BLOCKED is a hard stop for an unsupported extension, zero or oversized bytes,
supplied character/word limits, failed approval/consent/retention/destination
checks, unknown fields, duplicate keys, or unsafe content/credential fields. Do not
truncate, split, preview, log, or upload a document to investigate. Invalid input
returns a non-zero exit status and a JSON decision without the rejected value.
Examples
- A
.docx manifest at 4,194,304 bytes with approved classification, authority,
consent, transfer, retention, exact API origin, and exact inspected upload origin:
READY.
- The same content-bound manifest with
presigned_upload_origin: null and
presigned_upload_origin_approved: false: INSPECTION_READY, which permits only
destination discovery and no document upload.
- A
.txt manifest missing derived counts or with word_count: 29: BLOCKED.
- A
restricted manifest, lookalike API origin, private upload origin, or unapproved
inspected upload origin: BLOCKED.
- A manifest containing
preview, raw_text, or client_token: BLOCKED before
any size or governance decision is used.
Resources
- Read references/submission-contract.md for
the closed-world input schema and API-derived limits.
- Read references/governance-gate.md when deciding
what approval, consent, retention, and destination evidence is sufficient.
- The script is intentionally offline and has no write capability.
1---2name: grammarly-data-safety-guardian3description: Audit and gate a Grammarly document-submission manifest offline before an authorized integration upload. Use when checking supported extensions, file-size and text-length metadata, approval, consent, retention, and approved destination. This skill never accepts or submits document content and is not a replacement for legal, privacy, or security approval. Trigger with "Grammarly submission preflight", "Grammarly document safety", or "Grammarly retention check".4license: MIT5---6
7# Grammarly Data Safety Guardian
8
9Run a deterministic preflight over a metadata-only submission manifest. The guard
10answers whether a planned document upload satisfies this skill's hard safety gates;
11it does not read a document, contact Grammarly, create an upload URL, or retain a
12copy of the manifest.
13
14## Overview
15
16The guard is a closed-world policy check for a planned document submission. It
17combines Grammarly's documented format and size constraints with local governance
18decisions while keeping the document itself outside the workflow.
19
20## When to use
21
22Use for a Writing Score, AI Detection (Beta), or Plagiarism Detection (Beta)
23document submission before and after the evaluator inspects the provider-issued upload
24origin. The preliminary decision gates OAuth metadata disclosure; the final decision
25gates document bytes. Do not trigger for writing or editing content, document previews,
26credential troubleshooting, or a request to upload a real document.
27
28## Prerequisites
29
30- A synthetic or production-approved metadata-only manifest matching
31 [references/submission-contract.md](references/submission-contract.md).
32- Independent data-owner approval, consent evidence, transfer approval, content
33 classification, and acknowledgment of provider retention boundaries.
34- Python 3.10 or newer. No document path, document content, credential, network, or
35 writable output directory is required.
36
37## Instructions
38
391. Collect only the manifest described in
40 [references/submission-contract.md](references/submission-contract.md). The
41 manifest may contain extension, byte count, optional derived text counts, and
42 governance decisions, the pinned API control-plane origin, and the exact sanitized
43 presigned upload origin; it must not contain text, content, previews, credentials,
44 paths, filenames, full upload URLs, or raw API responses.
452. Run the guard with stdin:
46
47 ```bash
48 python3 scripts/audit_submission_manifest.py < submission-manifest.json
49 ```
50
513. Treat `BLOCKED` as a hard stop. Fix the named metadata or governance decision;
52 never truncate or transform a document silently to pass the guard.
534. Treat `INSPECTION_READY` as permission only to create a normalized-filename job and
54 inspect its sanitized upload origin. Treat `READY` as permission to continue to the separately authorized upload
55 workflow, not as proof that a document is lawful, accurate, or safe for a new
56 destination.
57
58The script accepts strict JSON, rejects duplicate keys and non-standard constants,
59recursively rejects raw text, location, response, and credential fields, and performs
60no network or filesystem writes. It enforces the documented extensions, 4,194,304-byte
61maximum, UTF-8 text counts, a non-restricted classification, all approval gates, the
62exact `https://api.grammarly.com` control-plane origin, and an explicitly approved
63public HTTPS S3 presigned-upload origin on the default TLS port.
64
65## Output
66
67Return the deterministic JSON decision and failed check names. Do not echo a
68filename, path, text, preview, credential, identifier, or raw document metadata
69beyond the safe numeric and enum fields in the script's output.
70
71## Error Handling
72
73`BLOCKED` is a hard stop for an unsupported extension, zero or oversized bytes,
74supplied character/word limits, failed approval/consent/retention/destination
75checks, unknown fields, duplicate keys, or unsafe content/credential fields. Do not
76truncate, split, preview, log, or upload a document to investigate. Invalid input
77returns a non-zero exit status and a JSON decision without the rejected value.
78
79## Examples
80
81- A `.docx` manifest at 4,194,304 bytes with approved classification, authority,
82 consent, transfer, retention, exact API origin, and exact inspected upload origin:
83 `READY`.
84- The same content-bound manifest with `presigned_upload_origin: null` and
85 `presigned_upload_origin_approved: false`: `INSPECTION_READY`, which permits only
86 destination discovery and no document upload.
87- A `.txt` manifest missing derived counts or with `word_count: 29`: `BLOCKED`.
88- A `restricted` manifest, lookalike API origin, private upload origin, or unapproved
89 inspected upload origin: `BLOCKED`.
90- A manifest containing `preview`, `raw_text`, or `client_token`: `BLOCKED` before
91 any size or governance decision is used.
92
93## Resources
94
95- Read [references/submission-contract.md](references/submission-contract.md) for
96 the closed-world input schema and API-derived limits.
97- Read [references/governance-gate.md](references/governance-gate.md) when deciding
98 what approval, consent, retention, and destination evidence is sufficient.
99- The script is intentionally offline and has no write capability.
gabrielmoreira