Hacktoberfest 2026: as issues que os mantenedores marcaram para outubro, abertas e boas para iniciantes. Ver issues do Hacktoberfest

FEAT: Cross-principal memory leakage scenario for agents with shared long-term memory

Aberta
  • Algo errado?
#3,272 0 comentários 0 reações 0 responsáveis Ver no GitHub

Mantenedores costumam responder em até 2 dias

Ninguém assumiu esta issue ainda.

Avaliação

Dificuldade
5/5
Tempo estimado
Mais de uma semana
Facilidade para iniciantes
18/100
Tipo de issue
Funcionalidade
Clareza
Razoavelmente clara
Status de atividade
Ativa
Stack de tecnologia
python
Domínio
security

Direção de pesquisa

The payload names the existing Leakage scenario and the substring/regex scorers as reference points, but no file paths. Start by reading how Leakage and those scorers are built, and ask the maintainers how multi-principal targets should be modelled before writing code. Done means a two-phase scenario that seeds canaries as one identity, probes as another, and reports leak rate per probe strategy.

Escrita pelo modelo de indexação a partir do texto da issue.

Descrição

feature-request
Is your feature request related to a problem? Please describe.

The Leakage scenario tests whether a model discloses training data, PII or confidential
information. It doesn't cover a growing agent-specific path: data that one user was entitled
to see is written into a shared long-term memory (e.g. a workspace-level store), and later
retrieved for a different user who is not entitled to it.

In this pattern, each individual check passes. The source query is authorized for user A,
the memory read is authorized by namespace or workspace membership for user B, and keyword
input/output filters pass because the second user's questions don't use restricted terms.
The leak exists only across sessions and principals, so single-conversation probing doesn't
surface it.

Two public descriptions of the pattern are in the OWASP GenAI Data Security Initiative's
trace dataset (memory_read_write category): an unintentional variant
(GenAI-Security-Project/GenAI-Data-Security-Initiative#102) and an adversarial variant where
a member deliberately words queries to extract restricted derived items (#104). Related
published work on query-only extraction from agent memory: Wang et al., "Unveiling Privacy
Risks in LLM Agent Memory" (arXiv:2502.13172).

Describe the solution you'd like

A scenario that runs in two phases against an agent target that supports per-user identity:

  1. Seed: as principal A, send prompts that cause the agent to read restricted data and
    persist something derived from it. Restricted values are unique canary tokens, so leakage
    can be detected deterministically.
  2. Probe: as principal B, who lacks the entitlement, send benign and adversarial
    (paraphrased, keyword-free, verbatim-recall) prompts drawn from a seed dataset.
  3. Score: a true/false scorer that flags any of A's canaries in B's responses. This could
    build on the existing substring/regex scorers, with no LLM judge needed.

Output: leak rate per probe strategy, which shows whether keyword filtering vs.
entitlement-aware memory reads actually stops the path.

Describe alternatives you've considered, if relevant
  • Extending Leakage directly. It's model-focused and single-principal, so a separate
    scenario (or a clearly separate technique set) seems cleaner, but I'm happy to follow your
    preference.
  • Shipping it as an external example notebook rather than a built-in scenario.
Additional context

I'd like to implement this if it fits PyRIT's direction, starting with the seed dataset and
the canary scorer, then the scenario. Guidance on how you'd want multi-principal targets
modelled would be very helpful before I start.

Disclosure: I contributed the two OWASP traces referenced above and have published on
lineage-aware access control for agent memory.

Linguagem predominante
Python
Estrelas
4.6k
Forks
944
Merge médio
2d 18h
PRs com merge (30d)
369

Preparar o ambiente

Abrir no Codespaces

Inicia o contêiner de desenvolvimento do projeto no navegador, com a sua própria conta do GitHub.

  • Sem Dockerfile nem arquivo Docker Compose
  • Tem um modelo de pull request
  • Sem guia de contribuição

Primeiros passos

  1. Leia a issue inteira e depois o guia de contribuição do projeto.
  2. Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
  3. Faça um fork do repositório e trabalhe em uma branch.
  4. Abra um pull request que referencie o número da issue.

Mais de microsoft/PyRIT

Todas as issues de microsoft/PyRIT

Issues semelhantes

Mais issues de Python

Receba novas issues na sua caixa de entrada

Um resumo curto de issues do GitHub para quem está começando.