Skip to content

[codex] follow up glm thinking recovery #63

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
terisuke merged 15 commits into from
Apr 5, 2026
Merged

[codex] follow up glm thinking recovery #63

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
8b826bb
fix: resolve upstream merge typecheck error in test harness (#22)
terisuke Apr 4, 2026
ae7eb4b
fix(guardrails): add bootstrap mode for HEAD-less repos (#25)
terisuke Apr 4, 2026
b06864c
fix(guardrails): remove OpenRouter eval-only restriction (#24)
terisuke Apr 4, 2026
d4a92b7
test(guardrails): add regression test for HEAD-less repo bootstrap mo…
terisuke Apr 4, 2026
552e30c
fix(ci): create JUnit output dirs before tests (#20959)
kitlangton Apr 4, 2026
aab1d29
feat(tui): add built-in system notifications for session completion
nianyi778 Apr 2, 2026
3d09150
fix(tui): improve notification error handling and escaping
nianyi778 Apr 2, 2026
38607a2
fix: use correct session.idle event type (sessionID not info.id)
nianyi778 Apr 2, 2026
667cd1c
fix(tui): harden notification delivery paths
terisuke Apr 4, 2026
e28ebc0
fix(team): allow fd-only bash redirection
terisuke Apr 5, 2026
bf1c1b9
Merge pull request #56
terisuke Apr 5, 2026
482833f
Merge pull request #59
terisuke Apr 5, 2026
5822a13
fix(ci): use pull_request_target for fork runner compatibility
terisuke Apr 5, 2026
b00d3d2
Merge pull request #64 from Cor-Incorporated/fix/ci-use-pull-request-…
terisuke Apr 5, 2026
d33e51a
follow up glm thinking recovery
terisuke Apr 5, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions .github/workflows/test.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ on:
push:
branches:
- dev
pull_request:
pull_request_target:
workflow_dispatch:

concurrency:
Expand All @@ -25,6 +25,7 @@ jobs:
- name: Checkout repository
uses: actions/checkout@v4
with:
ref: ${{ github.event.pull_request.head.sha || github.sha }}
token: ${{ secrets.GITHUB_TOKEN }}

- name: Setup Bun
Expand Down Expand Up @@ -61,7 +62,8 @@ jobs:
if: always()
uses: actions/upload-artifact@v4
with:
name: unit-linux-${{ github.run_attempt }}
name: unit-${{ matrix.settings.name }}-${{ github.run_attempt }}
Copy link

Copilot AI Apr 5, 2026

Choose a reason for hiding this comment

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

name: unit-${{ matrix.settings.name }}-... references matrix.* but this job does not define a matrix strategy. This will fail workflow evaluation at runtime/compile time; use a static name (e.g. unit-linux-...) or introduce an actual matrix.

Suggested change
name: unit-${{ matrix.settings.name }}-${{ github.run_attempt }}
name: unit-linux-${{ github.run_attempt }}

Copilot uses AI. Check for mistakes.
include-hidden-files: true
if-no-files-found: ignore
retention-days: 7
path: packages/*/.artifacts/unit/junit.xml
Expand All @@ -75,6 +77,7 @@ jobs:
- name: Checkout repository
uses: actions/checkout@v4
with:
ref: ${{ github.event.pull_request.head.sha || github.sha }}
token: ${{ secrets.GITHUB_TOKEN }}

- name: Setup Bun
Expand Down
4 changes: 3 additions & 1 deletion .github/workflows/typecheck.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: typecheck
on:
push:
branches: [dev]
pull_request:
pull_request_target:
branches: [dev]
workflow_dispatch:

Expand All @@ -13,6 +13,8 @@ jobs:
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
ref: ${{ github.event.pull_request.head.sha || github.sha }}

Comment on lines 3 to 18
Copy link

Copilot AI Apr 5, 2026

Choose a reason for hiding this comment

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

Switching to pull_request_target (and checking out the PR head SHA) runs untrusted PR code with the base-repo token context. For a typecheck workflow this is usually unnecessary risk; prefer pull_request or ensure the workflow never executes PR-provided code with elevated permissions/secrets.

Copilot uses AI. Check for mistakes.
- name: Setup Bun
uses: ./.github/actions/setup-bun
Expand Down
2 changes: 1 addition & 1 deletion packages/app/package.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@
"build": "vite build",
"serve": "vite preview",
"test": "bun run test:unit",
"test:ci": "bun test --preload ./happydom.ts ./src --reporter=junit --reporter-outfile=.artifacts/unit/junit.xml",
"test:ci": "mkdir -p .artifacts/unit && bun test --preload ./happydom.ts ./src --reporter=junit --reporter-outfile=.artifacts/unit/junit.xml",
"test:unit": "bun test --preload ./happydom.ts ./src",
"test:unit:watch": "bun test --watch --preload ./happydom.ts ./src",
"test:e2e": "playwright test",
Expand Down
2 changes: 1 addition & 1 deletion packages/guardrails/profile/AGENTS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,4 +10,4 @@
- Treat `.opencode/guardrails/` as plugin-owned runtime state, not a manual editing surface.
- Use `implement` as the guarded default primary agent. Route review, ship, and handoff work through the packaged `/review`, `/ship`, and `/handoff` commands instead of freeform release flows.
- Keep review paths read-only. If a workflow needs edits, return to `implement` or a project-local implementation agent instead of widening the review agent.
- Keep provider admission explicit. Standard confidential-code work stays on the admitted `zai` and `openai` lane; OpenRouter-backed evaluation belongs on `provider-eval` or `/provider-eval` only.
- All configured providers are available for standard work. The `provider-eval` agent and `/provider-eval` command remain available for dedicated evaluation workflows.
6 changes: 3 additions & 3 deletions packages/guardrails/profile/plugins/guardrail.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -202,7 +202,7 @@ export default async function guardrail(input: {
worktree: string
}, opts?: Record<string, unknown>) {
const mode = typeof opts?.mode === "string" ? opts.mode : "enforced"
const evals = new Set(["openrouter"])
const evals = new Set<string>([])
const evalAgent = "provider-eval"
const conf = true
const denyFree = true
Expand Down Expand Up @@ -349,10 +349,10 @@ export default async function guardrail(input: {
const agent = str(data.agent)
if (!provider) return

if (evals.has(provider) && agent !== evalAgent) {
if (evals.size > 0 && evals.has(provider) && agent !== evalAgent) {
return `${provider} is evaluation-only under confidential policy; use ${evalAgent}`
}
if (agent === evalAgent && !evals.has(provider)) {
if (evals.size > 0 && agent === evalAgent && !evals.has(provider)) {
return `${evalAgent} is reserved for evaluation-lane providers`
}

Expand Down
Loading
Loading