Skip to content

🔧🧑‍💻 - Make the Bugbot approval receipt producer trust boundary explicit and reproducible #980

Description

@efraespada

Is there an existing issue for this?

  • I have searched the existing issues.

What area does this task affect?

CI/CD

Task description

Make the Bugbot approval-receipt producer trust boundary explicit and reproducible. Deliver one independently reviewable result at the before any enforced activation stage.

Scope and acceptance criteria

  • Write the exact trust claim and prerequisite for any future enforced profile. Distinguish authenticated run/data provenance, trusted executable/workflow revision and operator producer attestation; do not call a matching artifact an independent code attestation.
  • Keep a runnable packaged regression for a fabricated clean current receipt with guarded policy and each changed path: bundle, workflow and action.yml. The decision must remain protected-path with zero approval writes, independently of any later stale-head fence.
  • At least eight consumer/packaged cases cover those three protected edits, a foreign run, foreign SHA, foreign App, prior policy/attempt and the ordinary valid authorized producer path. Preserve same-SHA policy change and rollback fences.
  • Document whether stronger producer evidence is required and any unavailable prerequisite. If architecture/public policy must change, use the official SDD review before implementation. This bounded audit must not invent or auto-enable an attestation system.
  • Keep the existing verified result, human decisions, read-only shadow pilots and production disabled. This issue does not reopen the refuted broad claim as a proven approval exploit.

Current issues or inefficiencies

Run/repository/SHA/artifact identity is not independent attestation of producer code. The reported action-edit approval route was not reproduced: guarded-policy packaged controls reject edits to build/**, the workflow and action.yml as protected-path. The narrower producer-trust limitation remains.

Observed baseline: unreleased develop ccefb27627cd627ccdb7d5666fd174c2ac1da07b from merged PR #951.

Expected impact

Make this one pending result independently inspectable and reproducible, with its actual passed, failed or unavailable evidence.
Completion does not authorize release, producer trust, profile qualification from partial measurements, or production activation.

Additional context or comments

Priority and dependencies

Documentation and implementation boundary

Documentation: docs/pull-requests/guarded-approval.mdx, docs/bugbot/quality-observability.mdx.
Keep Clean Architecture and current coverage/contract gates. When a later implementation changes a public contract, update its catalogued SDD, tests and docs together and validate specifications; update Graphify after code changes.
Read the effective #950 contract in the owning SDD and its current reconciliation amendments; historical pilot deadlines are not current investigation limits.

Fixed invariants and planning state

Backlog only: no in-progress, branch creation, model execution, deployment or activation requested by this planning step.
Keep BUGBOT_VERIFICATION_MODE=disabled. Evaluated PRs must have zero write attempts; retain frozen pilots and independent evaluator/oracle evidence.
Token usage is optional telemetry and cannot stop investigation or reject a supported verdict. Elapsed time and turn count do not impose research cutoffs; explicit cancellation and evidence/SHA integrity remain mandatory. Post-run latency qualification criteria remain distinct from session deadlines.

Activity

  1. moved this to Todo in vypdevon Nov 19, 2026
  2. changed the title [-]Make the Bugbot approval-receipt producer trust boundary explicit and reproducible[/-] [+]🔧 - Make the Bugbot approval receipt producer trust boundary explicit and reproducible[/+] on Nov 19, 2026
  3. vypbot commented on Nov 19, 2026

    @vypbot
    Contributor

    Hi! I’m @vypbot, the Copilot assistant for this repository.

    I can answer questions, explain the codebase, propose implementation and test plans, review issues and pull requests for potential bugs or security problems, and help authorized maintainers apply changes.

    Try /copilot help to see the available commands, or mention @vypbot with your question.

  4. moved this from Todo to In Progress in vypdevon Nov 19, 2026
  5. added
    branchedIndicates that a branch has been created for this issue.
    on Nov 19, 2026
  6. changed the title [-]🔧 - Make the Bugbot approval receipt producer trust boundary explicit and reproducible[/-] [+]🔧🧑‍💻 - Make the Bugbot approval receipt producer trust boundary explicit and reproducible[/+] on Nov 19, 2026
  7. added
    state:workingWork can proceed on the verified branch or without a branch.
    on Nov 19, 2026
  8. vypbot commented on Nov 19, 2026

    @vypbot
    Contributor

    Implementation plan

    Current status: Ready to start. No action is required from maintainers before implementation.

    1. Define the producer trust boundary
      • State separately what run/data provenance, trusted workflow and executable revision, and operator producer attestation establish.
      • Clarify that a matching artifact is not independent code attestation, and record whether stronger evidence is required and any unavailable prerequisite.
    2. Add reproducible packaged regressions
      • Exercise fabricated clean current receipts with guarded policy against changed bundle, workflow, and action.yml paths; each must return protected-path with zero approval writes, independent of the stale-head fence.
      • Cover at least eight consumer and packaged cases: those three edits, foreign run, foreign SHA, foreign App, prior policy or attempt, and the ordinary authorized producer path; retain same-SHA policy change and rollback fences.
    3. Document the bounded result
      • Update the guarded approval and Bugbot observability docs with the trust claim, prerequisite, evidence, and any unavailable prerequisite.
      • Keep the verified result, human decisions, read-only shadow pilots, and production-disabled state explicit; do not imply the broad approval-exploit claim was proven.
    4. Use contract review only if scope expands
      • If resolving the prerequisite requires a public policy or architecture change, update and review the owning SDD before implementation, then keep its contract, tests, and docs aligned.
      • Do not invent or automatically enable an attestation system; keep BUGBOT_VERIFICATION_MODE=disabled and evaluated PRs at zero write attempts.

    Acceptance: A reviewer can run the packaged regression and verify all required cases, including protected-path outcomes with zero approval writes for each protected edit; the trust-boundary documentation distinguishes provenance, trusted revision, and operator attestation, records any unavailable prerequisite, and confirms production remains disabled.

    Need something else? Mention the bot with a question or use /copilot help.

  9. added
    size: XXLIndicates an extremely large task or issue that requires significant time and effort to complete.
    75%Progress: 75%
    and removed on Nov 19, 2026
  10. vypbot commented on Nov 19, 2026

    @vypbot
    Contributor

    Progress · 75% estimated

    Assessed head 6efac734 · 6/10 supported
    0 need implementation · 4 need verification or clarification

    Verification or clarification needed

    • R11 · Make the Bugbot approval-receipt producer trust boundary explicit and reproducible. Deliver one independently reviewable result at the before any enforced activation stage. — Please independently verify the reported packaged results and publication counter-control. Requirement source · Evidence at assessed head
    • R34 · Make this one pending result independently inspectable and reproducible, with its actual passed, failed or unavailable evidence. — Please independently verify the reported passed, failed, or unavailable evidence for this result. Requirement source · Evidence at assessed head
    • R55 · Keep BUGBOT_VERIFICATION_MODE=disabled. Evaluated PRs must have zero write attempts; retain frozen pilots and independent evaluator/oracle evidence. — Please provide independent confirmation that production stayed disabled, pilots and evaluator evidence were retained, and evaluated PRs had zero write attempts. Requirement source · Evidence at assessed head
    • R56 · Token usage is optional telemetry and cannot stop investigation or reject a supported verdict. Elapsed time and turn count do not impose research cutoffs; explicit cancellation and — Please clarify whether evidence verifies these process constraints outside this bounded audit. Requirement source · Evidence at assessed head

    Issue · Linked pull request · Full assessment

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

75%Progress: 75%branchedIndicates that a branch has been created for this issue.chorein-progressStart work on an admitted issue.maintenancepriority: mediumImportant but not urgent, should be addressed soon.size: XXLIndicates an extremely large task or issue that requires significant time and effort to complete.state:workingWork can proceed on the verified branch or without a branch.

Projects

  • Status
    In Progress

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions