Open
Conversation
Author
|
@microsoft-github-policy-service agree |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Implemented a Word document file converter to support indirect prompt injection scenarios, addressing FEAT File Converter: Word Document (#424).
Feature: WordDoc_Converter
Added WordDoc_Converter in pyrit.prompt_converter.word_doc_converter as a PromptConverter that transforms text prompts into .docx files.
Exposed via pyrit.prompt_converter.init for discovery through get_converter_modalities.
Supported Modes
New Document Generation
Creates a simple .docx with optionally templated prompt content in a single paragraph.
Placeholder-Based Injection
Replaces a literal placeholder in an existing .docx (within paragraph runs) with rendered content while preserving formatting.
Templating and Safety
Uses SeedPrompt.render_template_value for templating.
Does not render Jinja2 templates from arbitrary .docx content.
Injected content is plain text output from SeedPrompt.
Formatting Constraints
Placeholders must be fully contained within a single run.
If spanning multiple runs, replacement is skipped and a warning is logged to avoid formatting issues.
Tests
Added unit tests for:
Creating a new .docx from a plain text prompt.
Injecting content when the placeholder is within a single run.
Verifying no replacement and warning when placeholders span multiple runs.
Documentation
Updated documentation to describe:
WordDoc_Converter
Supported modes
Placeholder requirements
Use of SeedPrompt for templating