Add RAG-0001 synthetic poisoning regression fixture - #58
Merged
emmanuelgjr merged 1 commit intoSep 14, 2026
Merged
emmanuelgjr merged 1 commit into
emmanuelgjr merged 1 commit into
Conversation
Adds RAG-0001.json for ingestion and retrieval integrity testing. Signed-off-by: Subhash Dasyam <subhashdasyam@users.noreply.github.com>
There was a problem hiding this comment.
Pull request overview
Adds a first, fully synthetic “poisoned document” fixture (RAG-0001) to the RAG poisoning & retrieval integrity dataset, intended to support regression testing of ingestion quarantine, authorized retrieval filtering, and instruction-resilience behavior in RAG pipelines.
Changes:
- Adds a new dataset entry (
RAG-0001) describing a two-stage synthetic poisoning scenario targeting a return-policy query. - Includes secure vs. vulnerable expected outcomes plus evaluation flags/canary strings and supporting reference links.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
emmanuelgjr
approved these changes
Sep 14, 2026
emmanuelgjr
left a comment
emmanuelgjr
left a comment
Contributor
There was a problem hiding this comment.
Validated RAG-0001 against rag.schema.json; synthetic content, DSGAI04/05/13 mapping matches the dataset README, references are primary sources. Thanks.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Adds the first synthetic entry to the RAG poisoning and retrieval integrity dataset.
RAG-0001is a safe two-stage test for:The scenario uses invented policy data, invokes no tools, includes no real identifiers, and requests no external action. The entry follows the repository's current
rag.schema.jsonand one-entry-per-file rule.I developed the test while updating a detailed practitioner guide to RAG poisoning controls and negative tests: https://www.subhashdasyam.com/2024/09/vector-venom-vector-injection-how-ai.html
Primary research and guidance are included in the entry metadata.
DSGAI mapping
Validation
data_validation/schemas/rag.schema.jsonpassesChecklist