Skip to content

fix: add anti-preamble instructions to role play attack system prompts - #2900

Open
ANIRUDDHA ADAK (aniruddhaadak80) wants to merge 4 commits into
microsoft:mainfrom
aniruddhaadak80:fix/role-play-no-preamble
Open

ANIRUDDHA ADAK (aniruddhaadak80) wants to merge 4 commits into
microsoft:mainfrom
aniruddhaadak80:fix/role-play-no-preamble

Conversation

@aniruddhaadak80

Copy link
Copy Markdown

Summary

Fixes #1912 — role play attacks were sometimes starting responses with conversational preamble like "Sure, here is your objective rephrased as a detailed question set..." instead of directly outputting the attack prompt.

Changes

  • Added explicit anti-preamble instructions to all 5 role play persona YAMLs (role_play_movie_script, role_play_video_game, role_play_trivia_game, role_play_persuasion, role_play_persuasion_written)
  • Added the same instruction to role_play_next_message.yaml
  • Added tests verifying the anti-preamble instruction is present in all role play YAMLs

Testing

  • Added test_yaml_has_no_preamble_instruction parametrized test for all 5 role play persona YAMLs
  • Added test_role_play_next_message_has_no_preamble_instruction for the next-message prompt
  • All YAML files verified to load correctly and contain the new instructions

Comment thread tests/unit/setup/test_technique_initializer.py
"""Role-play persona YAMLs must instruct the model to avoid preamble/filler."""
path = Path(EXECUTOR_SEED_PROMPT_PATH) / "red_teaming" / "role_play" / f"{technique_name}.yaml"
text = path.read_text(encoding="utf-8")
assert "must begin directly with the in-character attack prompt" in text, (

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟡 Should fix: These assertions only prove the YAML contains the new words. They cannot show whether the generated next_message actually loses its preamble, which is the behavior requested in #1912. The adversarial reply parser accepts any nonempty next_message, including "Sure, here is...", and forwards it unchanged. Could you include the issue's before/after 50-run comparison, counting preambles in the final target-directed message, or other measured output evidence? A green substring test alone does not establish that the fix works.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed, I don't want this PR to strip or rewrite generated messages. The parser should preserve the model's output, and adding a content filter would be a separate change. Sorry, mentioning the parser made my request less clear.

My request was for measured before/after evidence that the prompt change reduces preambles, as described in #1912. The substring tests are useful guards for the instruction, but they cannot establish its effect on generated output. Could you run the 50-before/50-after comparison using the same model and settings, and count preambles in the final target-directed next_message? Please include the model/settings and report refusals or errors separately. Benign objectives are fine for this comparison; no parser changes or live-model unit tests are needed.

Auto-replied by the GitHub Copilot app

@aniruddhaadak80
ANIRUDDHA ADAK (aniruddhaadak80) force-pushed the fix/role-play-no-preamble branch 2 times, most recently from 141a633 to 205e154 Compare September 30, 2026 09:54
@aniruddhaadak80

Copy link
Copy Markdown
Author

Done in bf7a5961 and 205e154f — the must-fix is in, plus a second defect I found while verifying it. Also rebased onto main (was 7 behind).

🔴 Missing import. Your read was right and it reproduced exactly:

FAILED tests/unit/setup/test_technique_initializer.py::TestRolePlayYamls::test_role_play_next_message_has_no_preamble_instruction
E       NameError: name 'EXECUTOR_SIMULATED_TARGET_PATH' is not defined
1 failed, 22 passed

EXECUTOR_SIMULATED_TARGET_PATH is defined at pyrit/common/path.py:61; line 11 of the test imported only the other two. Added it, so that thread is closed.

A second one I hadn't caught. Checking the diff against main rather than just running the test, the anti-preamble bullet had also re-indented neighbouring lines one column to the right in all six prompts:

  • ## Response Format went from two spaces to three in the five personas
  • a run of bullets under ## Hard Constraints in role_play_next_message.yaml went from two to three, leaving that one list mixed 2/3/2

YAML strips a block scalar's base indent, so this isn't cosmetic. In the next-message prompt the stray column reaches the model and renders the flat list as a nested sub-list, and the heading moves out of column zero. Only the first content line sets the base indent, so the files still parsed, still built, and nothing complained. That is exactly why it survived review. Fixed in bf7a5961 — every pre-existing line is back where it was, and the net diff against main is now purely additive: +99/-1 across 7 files, where before it was +55/-8 with 6 deletions that were never mine to make.

I added a guard in 205e154f over all six prompts (not just the five personas), asserting every ## heading and top-level bullet inside a value: | block sits at the block's base indent. I confirmed it fails on the pre-fix indentation and passes after, so it is a real regression test and not a tautology.

🟡 The behavioural assertion — I'd like to push back rather than silently rework this.

You are right that the assertions only prove the YAML contains the words. But the parser accepting a nonempty next_message including "Sure, here is..." is not a gap in this PR, it is the design. #1912 asked for the instruction; the anti-preamble behaviour is produced by the model obeying the prompt, and the reply parser is the boundary where the model's JSON is read, not a content filter. Having _AdversarialConversationManager defensively rewrite a nonempty next_message would be a separate design change on a different code path — it would also have to decide where the preamble ends, which for "Let me..." vs "Let me explain X" is a judgement the parser has no basis to make, and it would silently alter model output in a red-teaming harness where fidelity to the transcript is the point.

If you do want that hardening, I think it belongs in its own PR against _generate_next_message_async, and I'd rather write it with you than smuggle it in behind a prompt change. Happy to be told I'm wrong and to write it.

ruff check and ruff format --check are clean, and the full file is 109 passed / 0 failed.

One thing still outstanding, and it is not mine to do: the CLA is unsigned, so license/cla is pending and the PR stays blocked. That is the microsoft-github-policy-service agreement, which needs you.

Role play attacks were sometimes starting responses with conversational
preamble like 'Sure, here is your objective rephrased as a detailed
question set...' instead of directly outputting the attack prompt.

Added explicit instructions to all 5 role play persona YAMLs and the
role_play_next_message.yaml to begin next_message directly with the
in-character attack prompt, with no preamble, conversational filler,
or meta commentary.

Fixes microsoft#1912

Signed-off-by: Aniruddha Adak <aniruddhaadak80@users.noreply.github.com>
Adding the anti-preamble bullet re-indented neighbouring lines one column to
the right in all six prompts: ## Response Format went from two spaces to
three in the five personas, and a run of bullets under Hard Constraints in
role_play_next_message.yaml went from two to three.

YAML strips a block scalar's base indent, so the stray column is not
cosmetic. It reaches the model: the next-message prompt's flat bullet list
became a 2/3/2 mix, which renders as a nested sub-list, and the section
heading moved out of column zero. Only the first content line sets the base
indent, so the files still parse and still build, which is why nothing
flagged this.

Every pre-existing line is back where it was; the net diff against main is
now purely additive.
…tion

The next-message test raised NameError because EXECUTOR_SIMULATED_TARGET_PATH
was never imported; it is defined in pyrit.common.path. Added it alongside the
other two executor paths. Verified: 23 passed on -k role_play before this, 1
failed with NameError on the branch.

Also guards the prompt indentation regression fixed in the previous commit.
Every ##  heading and top-level bullet inside a �alue: | block must sit at
the block's base indent, across all six role-play prompts rather than the five
personas alone. Confirmed the test fails on the pre-fix indentation and passes
after.

The behavioural point from the review is deliberately not addressed here: the
parser does not and should not strip preamble. This change is a prompt
instruction, and the instruction is what the model is asked to obey. Making the
reply parser defensively rewrite a nonempty next_message would be a different
design change, on a different code path, and out of scope for a prompt fix.
Copilot AI balanced review requested due to automatic review settings October 3, 2026 21:43

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@romanlutz Roman Lutz (romanlutz) self-assigned this Oct 9, 2026
@microsoft-github-policy-service

Copy link
Copy Markdown

ANIRUDDHA ADAK (@aniruddhaadak80) please read the following Contributor License Agreement(CLA). If you agree with the CLA, please reply with the following information.

@microsoft-github-policy-service agree [company="{your company}"]

Options:

  • (default - no company specified) I have sole ownership of intellectual property rights to my Submissions and I am not making Submissions in the course of work for my employer.
@microsoft-github-policy-service agree
  • (when company given) I am making Submissions in the course of work for my employer (or my employer has intellectual property rights in my Submissions by contract or applicable law). I have permission from my employer to make Submissions and enter into this Agreement on behalf of my employer. By signing below, the defined term “You” includes me and my employer.
@microsoft-github-policy-service agree company="Microsoft"
Contributor License Agreement

Contribution License Agreement

This Contribution License Agreement (“Agreement”) is agreed to by the party signing below (“You”),
and conveys certain license rights to Microsoft Corporation and its affiliates (“Microsoft”) for Your
contributions to Microsoft open source projects. This Agreement is effective as of the latest signature
date below.

  1. Definitions.
    “Code” means the computer software code, whether in human-readable or machine-executable form,
    that is delivered by You to Microsoft under this Agreement.
    “Project” means any of the projects owned or managed by Microsoft and offered under a license
    approved by the Open Source Initiative (www.opensource.org).
    “Submit” is the act of uploading, submitting, transmitting, or distributing code or other content to any
    Project, including but not limited to communication on electronic mailing lists, source code control
    systems, and issue tracking systems that are managed by, or on behalf of, the Project for the purpose of
    discussing and improving that Project, but excluding communication that is conspicuously marked or
    otherwise designated in writing by You as “Not a Submission.”
    “Submission” means the Code and any other copyrightable material Submitted by You, including any
    associated comments and documentation.
  2. Your Submission. You must agree to the terms of this Agreement before making a Submission to any
    Project. This Agreement covers any and all Submissions that You, now or in the future (except as
    described in Section 4 below), Submit to any Project.
  3. Originality of Work. You represent that each of Your Submissions is entirely Your original work.
    Should You wish to Submit materials that are not Your original work, You may Submit them separately
    to the Project if You (a) retain all copyright and license information that was in the materials as You
    received them, (b) in the description accompanying Your Submission, include the phrase “Submission
    containing materials of a third party:” followed by the names of the third party and any licenses or other
    restrictions of which You are aware, and (c) follow any other instructions in the Project’s written
    guidelines concerning Submissions.
  4. Your Employer. References to “employer” in this Agreement include Your employer or anyone else
    for whom You are acting in making Your Submission, e.g. as a contractor, vendor, or agent. If Your
    Submission is made in the course of Your work for an employer or Your employer has intellectual
    property rights in Your Submission by contract or applicable law, You must secure permission from Your
    employer to make the Submission before signing this Agreement. In that case, the term “You” in this
    Agreement will refer to You and the employer collectively. If You change employers in the future and
    desire to Submit additional Submissions for the new employer, then You agree to sign a new Agreement
    and secure permission from the new employer before Submitting those Submissions.
  5. Licenses.
  • Copyright License. You grant Microsoft, and those who receive the Submission directly or
    indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license in the
    Submission to reproduce, prepare derivative works of, publicly display, publicly perform, and distribute
    the Submission and such derivative works, and to sublicense any or all of the foregoing rights to third
    parties.
  • Patent License. You grant Microsoft, and those who receive the Submission directly or
    indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license under
    Your patent claims that are necessarily infringed by the Submission or the combination of the
    Submission with the Project to which it was Submitted to make, have made, use, offer to sell, sell and
    import or otherwise dispose of the Submission alone or with the Project.
  • Other Rights Reserved. Each party reserves all rights not expressly granted in this Agreement.
    No additional licenses or rights whatsoever (including, without limitation, any implied licenses) are
    granted by implication, exhaustion, estoppel or otherwise.
  1. Representations and Warranties. You represent that You are legally entitled to grant the above
    licenses. You represent that each of Your Submissions is entirely Your original work (except as You may
    have disclosed under Section 3). You represent that You have secured permission from Your employer to
    make the Submission in cases where Your Submission is made in the course of Your work for Your
    employer or Your employer has intellectual property rights in Your Submission by contract or applicable
    law. If You are signing this Agreement on behalf of Your employer, You represent and warrant that You
    have the necessary authority to bind the listed employer to the obligations contained in this Agreement.
    You are not expected to provide support for Your Submission, unless You choose to do so. UNLESS
    REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING, AND EXCEPT FOR THE WARRANTIES
    EXPRESSLY STATED IN SECTIONS 3, 4, AND 6, THE SUBMISSION PROVIDED UNDER THIS AGREEMENT IS
    PROVIDED WITHOUT WARRANTY OF ANY KIND, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTY OF
    NONINFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
  2. Notice to Microsoft. You agree to notify Microsoft in writing of any facts or circumstances of which
    You later become aware that would make Your representations in this Agreement inaccurate in any
    respect.
  3. Information about Submissions. You agree that contributions to Projects and information about
    contributions may be maintained indefinitely and disclosed publicly, including Your name and other
    information that You submit with Your Submission.
  4. Governing Law/Jurisdiction. This Agreement is governed by the laws of the State of Washington, and
    the parties consent to exclusive jurisdiction and venue in the federal courts sitting in King County,
    Washington, unless no federal subject matter jurisdiction exists, in which case the parties consent to
    exclusive jurisdiction and venue in the Superior Court of King County, Washington. The parties waive all
    defenses of lack of personal jurisdiction and forum non-conveniens.
  5. Entire Agreement/Assignment. This Agreement is the entire agreement between the parties, and
    supersedes any and all prior agreements, understandings or communications, written or oral, between
    the parties relating to the subject matter hereof. This Agreement may be assigned by Microsoft.

Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Comment on lines +500 to +501
assert "–" not in text, f"{technique_name}.yaml contains an en-dash"
assert "—" not in text, f"{technique_name}.yaml contains an em-dash"

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟡 Should Fix: These edits change the existing dash checks into checks for mojibake, so they no longer reject an actual en dash or em dash. I called both TestRolePlayYamls.test_yaml_has_no_em_or_en_dashes and TestPersonaCrescendoYamls.test_yaml_has_no_em_or_en_dashes with file contents containing U+2013 or U+2014, and all four checks passed.

Please restore the original needles in both classes. ASCII escapes also avoid an encoding round-trip:

assert "\u2013" not in text
assert "\u2014" not in text

This change is unrelated to the preamble fix and weakens an existing regression check.

This branch has not been deployed

No deployments
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.

BUG role play attack should respond with attack prompt only

3 participants