Declare 25 audited read-only actions across eight agents - #3042
Merged
George Ng (GeorgeNgMsft) merged 3 commits intoSep 21, 2026
Merged
Conversation
Cover eight actions across five agents with production-manifest and structured-service regression tests and an action rationale table. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Use fixed Windows ipconfig help and configuration reads instead; preserve the eight-action rationale table and production contract coverage. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Keep the rationale table in local review documents, exercise real handlers with isolated storage and mocked transport, and treat bare GitHub repository queries as literal arguments. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
George Ng (GeorgeNgMsft)
marked this pull request as ready for review
September 21, 2026 07:33
George Ng (GeorgeNgMsft)
enabled auto-merge
September 21, 2026 07:33
Hillary Mutisya (hillary-mutisya)
approved these changes
Sep 21, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Allow 25 audited read actions across eight built-in agents to skip the structured dispatcher's outer effect-confirmation prompt. This expands the original eight-action pass while keeping authorization, readiness, validation, explicit confirmation requirements, and handler questions intact. Natural-language routing and dispatcher defaults are unchanged.
{ "effects": "read-only" }policies for list inventory/content (2), weather conditions/forecast (2), Windows ipconfig help/configuration/DNS-cache display (3), GitHub read queries (12), reminder inventory (1), local-player status/queue/folder (3), PowerShell flow inventory (1), and task-flow inventory (1).--so flag-like names such as--webremain literal input. This narrowly scoped handler fix is necessary for safe repository-aware exemptions; repository-choice prompts are preserved.Validation: affected GitHub/provider builds passed; 49 production-policy regression tests, 90 shared structured discovery/execution tests, and 158 existing GitHub argument/readiness/diagnostic/result tests passed. Pinned changed-file formatting and all four committed-diff ratchets against
origin/mainpassed: no new lint violations, complexity, circular dependencies, focused tests, or skipped tests. No live external mutations, authentication, playback, or GUI actions were performed.Scope remains intentionally narrow: Spotify reads can invoke interactive auth; the calendar date-range adapter needs separate work; unrestricted utility/file/script execution, GitHub token exposure and annotation-host fetching, playback controls, and mutations remain unclassified and confirmed. ipconfig is Windows-only. Only flow metadata inventories are exempted, not scripts or PowerShell namespace actions. Existing lifecycle initialization and background behavior are unchanged.
This enables unattended read scenarios, not a universal speedup claim: discovery, connection, model time, and warm NL cache hits remain separate costs. The benchmarked list policies and handlers are unchanged.