Disable default Copilot sync assignment

Author: KSemenenko

.github/workflows/codex-cli-watch.yml

@@ -218,7 +218,7 @@ jobs:
           CHANGED_FLAGS: ${{ steps.detect.outputs.changed_flags }}
           CHANGED_MODELS: ${{ steps.detect.outputs.changed_models }}
           CHANGED_FEATURES: ${{ steps.detect.outputs.changed_features }}
-          ASSIGNEE: ${{ vars.SDK_SYNC_ASSIGNEE != '' && vars.SDK_SYNC_ASSIGNEE || 'copilot' }}
+          ASSIGNEE: ${{ vars.SDK_SYNC_ASSIGNEE }}
         with:
           script: |
             const currentSha = process.env.CURRENT_SHA;
@@ -229,7 +229,7 @@ jobs:
             const changedFlags = process.env.CHANGED_FLAGS || '- (not provided)';
             const changedModels = process.env.CHANGED_MODELS || '- (not provided)';
             const changedFeatures = process.env.CHANGED_FEATURES || '- (not provided)';
-            const assignee = process.env.ASSIGNEE || 'copilot';
+            const assignee = (process.env.ASSIGNEE || '').trim();
 
             const shortCurrent = currentSha.slice(0, 7);
             const shortLatest = latestSha.slice(0, 7);
@@ -332,50 +332,65 @@ jobs:
                 labels: [labelName],
               });
 
-              try {
-                await github.rest.issues.addAssignees({
-                  owner: context.repo.owner,
-                  repo: context.repo.repo,
-                  issue_number: reusableIssue.number,
-                  assignees: [assignee],
-                });
-                core.info(`Updated existing sync issue #${reusableIssue.number} and assigned @${assignee}`);
-              } catch (error) {
-                core.warning(`Issue assignment failed (${error.message}) for #${reusableIssue.number}.`);
-                await github.rest.issues.createComment({
-                  owner: context.repo.owner,
-                  repo: context.repo.repo,
-                  issue_number: reusableIssue.number,
-                  body: `Could not auto-assign @${assignee}. Please assign manually.`,
-                });
+              if (assignee) {
+                try {
+                  await github.rest.issues.addAssignees({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    issue_number: reusableIssue.number,
+                    assignees: [assignee],
+                  });
+                  core.info(`Updated existing sync issue #${reusableIssue.number} and assigned @${assignee}`);
+                } catch (error) {
+                  core.warning(`Issue assignment failed (${error.message}) for #${reusableIssue.number}.`);
+                  await github.rest.issues.createComment({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    issue_number: reusableIssue.number,
+                    body: `Could not auto-assign @${assignee}. Please assign manually.`,
+                  });
+                }
+              } else {
+                core.info(`Updated existing sync issue #${reusableIssue.number} without assignee.`);
               }
             } else {
-              try {
+              if (assignee) {
+                try {
+                  issue = await github.rest.issues.create({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    title,
+                    body,
+                    labels: [labelName],
+                    assignees: [assignee],
+                  });
+                  core.info(`Created and assigned issue #${issue.data.number} to @${assignee}`);
+                } catch (error) {
+                  core.warning(`Issue assignment failed (${error.message}), creating without assignee.`);
+                  issue = await github.rest.issues.create({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    title,
+                    body,
+                    labels: [labelName],
+                  });
+
+                  await github.rest.issues.createComment({
+                    owner: context.repo.owner,
+                    repo: context.repo.repo,
+                    issue_number: issue.data.number,
+                    body: `Could not auto-assign @${assignee}. Please assign manually.`,
+                  });
+                }
+              } else {
                 issue = await github.rest.issues.create({
                   owner: context.repo.owner,
                   repo: context.repo.repo,
                   title,
                   body,
                   labels: [labelName],
-                  assignees: [assignee],
-                });
-                core.info(`Created and assigned issue #${issue.data.number} to @${assignee}`);
-              } catch (error) {
-                core.warning(`Issue assignment failed (${error.message}), creating without assignee.`);
-                issue = await github.rest.issues.create({
-                  owner: context.repo.owner,
-                  repo: context.repo.repo,
-                  title,
-                  body,
-                  labels: [labelName],
-                });
-
-                await github.rest.issues.createComment({
-                  owner: context.repo.owner,
-                  repo: context.repo.repo,
-                  issue_number: issue.data.number,
-                  body: `Could not auto-assign @${assignee}. Please assign manually.`,
                 });
+                core.info(`Created sync issue #${issue.data.number} without assignee.`);
               }
             }
 

AGENTS.md

@@ -89,7 +89,7 @@ If no new rule is detected -> do not update the file.
 - When asked to fix review findings, close every confirmed finding in the same pass; do not leave partial fixes.
 - Do not keep or add public sample projects; repository focus is SDK + tests only.
 - Upstream sync automation must track real `openai/codex` CLI changes (flags/models/features), not TypeScript SDK surface diffs, and open actionable repository issues for required SDK follow-up.
-- Automatically opened upstream sync issues must include change summary/checklist and assign Copilot by default.
+- Automatically opened upstream sync issues must include change summary/checklist and stay unassigned unless the user explicitly requests an assignee.
 - For `openai/codex` repo sync/update work, always inspect the bundled `models.json` catalog in `submodules/openai-codex` (prefer `codex-rs/models-manager/models.json`, fall back to `codex-rs/core/models.json` for older pins) and reconcile SDK model constants against that repo-authoritative source.
 - At the end of implementation/code-change tasks, create a git commit unless the user explicitly says not to, so the workspace ends in a reviewable state.
 - Run verification in this order:

README.md

@@ -99,7 +99,7 @@ var thread = client.StartThread(new ThreadOptions
 
 `Ephemeral = false` forces rollout persistence even if your Codex config/profile enables `ephemeral = true`.
 
-SDK turns still run through non-interactive `codex exec`, so the default CLI/App history views may hide them. To reopen a persisted SDK session outside the SDK, use `codex resume <threadId>` directly or `codex resume --include-non-interactive`.
+SDK turns still run through non-interactive `codex exec`, so the default CLI/App history views may hide them. To reopen a persisted SDK session outside the SDK, use `codex resume <threadId>` directly.
 
 ## Codex CLI Metadata
 

docs/Features/release-and-sync-automation.md

@@ -47,7 +47,7 @@ Keep package quality and upstream Codex CLI parity automatically verified throug
 - Completing a Codex CLI sync issue must update the pinned `submodules/openai-codex` commit after validation.
 - Sync issue body must derive flag changes from CLI source snapshots, model changes from the bundled `models.json` catalog (`codex-rs/models-manager/models.json` in current upstream, with fallback to `codex-rs/core/models.json` for older pins), and feature changes from `codex-rs/core/config.schema.json` so alerts stay actionable.
 - SDK model constants must cover every bundled slug from the pinned `submodules/openai-codex` `models.json` catalog whenever upstream Codex repo sync work updates the pinned submodule.
-- Sync issue must assign Copilot by default.
+- Sync issue must stay unassigned by default unless repository configuration explicitly provides an assignee.
 - Sync issue automation must keep at most one open `codex-cli-sync` issue at a time by updating the active issue and closing superseded ones when upstream advances again.
 
 ---

docs/Features/thread-run-flow.md

@@ -47,7 +47,7 @@ Provide deterministic thread-based execution over Codex CLI so C# consumers can
 - Codex executable resolution is deterministic: prefer npm-vendored native binary, then PATH lookup; on Windows PATH lookup checks `codex.exe`, `codex.cmd`, `codex.bat`, then `codex`.
 - Thread options map full Codex CLI flags (`profile`, `enable/disable`, OSS provider, ephemeral/color/progress/output options), plus raw `AdditionalCliArguments` passthrough for forward-compatible flags.
 - Explicit `ThreadOptions.Ephemeral = false` must override inherited Codex config/profile `ephemeral = true`, so SDK callers can force persistent rollout storage deterministically.
-- Non-ephemeral SDK runs persist as non-interactive `codex exec` sessions; they remain resumable by thread id, but Codex CLI pickers may require `codex resume --include-non-interactive` to list them.
+- Non-ephemeral SDK runs persist as non-interactive `codex exec` sessions; they remain resumable by thread id even when default CLI/App history pickers do not show them.
 - If thread web search options are not set, SDK does not emit `web_search` overrides and keeps effective CLI/config setting unchanged.
 - Cleanup failures are never silently swallowed; process/schema/image cleanup issues are logged through `ILogger`.