[juliusmarminge]

In the OpenCode v2 adapter, prepareTurn already left out the orchestration guidance when T3's MCP tools were not registered (an external Server URL, or a failed mcp.add). It still called buildRuntimeInstructions without pullRequestLinking, so the <pull_request_linking> block telling the agent to call link_pull_request, watch_pull_request and unwatch_pull_request went into the session instructions on every turn even though those tools did not exist. The fix passes pullRequestLinking: mcpAvailable, the same flag the orchestration guidance uses. The ACP adapter passes this option the same way. One call site changed, no new parser or regex. The two tests on factory/repro-35588 (external server, and MCP add failed) fail without the fix and pass with it. The bug was reproduced with those tests, not against a live external OpenCode server.

Fixes #35588

Reproduction on main

Commit 932cbd4 (origin/main). Started OpenCode 2.2.7 (@opencode/cli) as an independent server: opencode serve --hostname 127.0.0.1 --port 49374 with isolated XDG config/data dirs. Started T3 with T3CODE_DEV_SEED=0 vp run dev and paired the preview browser. In Settings > Providers > Add provider I added an OpenCode instance and set Binary path, Server URL http://127.0.0.1:49374 and the server password. In a new thread I chose Claude Haiku 5.5 on that provider and sent one message asking the agent to list its tools whose names contain pull_request. The turn completed. On the OpenCode server, GET /api/mcp?location[directory]=<thread dir> returned data: [], so no T3 MCP was registered. GET /api/experimental/session/<id>/instructions/entries returned key t3-code, containing <runtime_info> and the full <pull_request_linking> block (link_pull_request, watch_pull_request, unwatch_pull_request, 'no sleep, loops ... for PR waits') and no orchestration guidance. The agent replied: "I don't have any tools with "pull_request" in their name, so the list is empty." On branch factory/repro-35588 (5a23f69), two tests in OpenCode2AdapterV2.test.ts fail on main with an instructions-frame mismatch: the external-server case and the existing MCP-add-failed case. Passing pullRequestLinking: mcpAvailable makes all 158 tests in the file pass; that fix is not on the branch.

Evidence

Evidence: configure an external OpenCode 2.2.7 provider on each fresh T3 server, run the same tool-list request, then inspect the completed conversation and OpenCode's session instructions. Both sessions have an empty MCP list. Base includes mandatory PR-tool instructions; head omits them while retaining runtime information.

Before	After
932cbd4e44	857fff9696

Transcript images render actual endpoint output, with runtime attribution redacted. Web captures use the same viewport, theme and crop. Before transcript · After transcript · Reproduction driver · Instruction capture script

AI-assisted; verified with the tests and tools listed above. No tests were used as proof.

Not shown:

• Failed MCP registration on a T3-managed OpenCode server was not exercised; this proof covers the external Server URL path.

Tests

vp test run src/orchestration-v2/Adapters/OpenCode2AdapterV2.test.ts (apps/server): without the fix, 2 failed and 156 passed (the two repro tests); with the fix, 158 passed. vp run check:changed: format, lint, knip, graphql, typecheck (10 packages) and tests (2 files, 160 tests) all passed.

AI-assisted; verified with the tests and tools listed above.

[macroscopeapp]

Approvability

Verdict: Approved at 857fff9

Macroscope's review found this PR approvable — OpenCode sessions without T3 MCP will no longer receive instructions to invoke unavailable PR tools. Sessions with MCP retain their existing guidance, keeping the change narrowly scoped.

No code changes detected at ab693a8. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	21.5 KiB	21.5 KiB	0 B (0.0%)	24.4 KiB	✅
Codex	Live turn messages	1	1	0 (0.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.9 KiB	21.9 KiB	0 B (0.0%)	24.4 KiB	✅
Claude	Live turn messages	1	1	0 (0.0%)	8	✅

Other scenarios

All 64 scenario measurements match main.

All 64 scenario measurements

Provider	Scenario	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	short turn resume	WebSocket decoded	10.5 KiB	10.5 KiB	0 B (0.0%)	12.7 KiB	✅
Codex	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn running, cold open	HTTP decoded	281.7 KiB	281.7 KiB	0 B (0.0%)	302.7 KiB	✅
Codex	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Codex	long turn running, history page	HTTP decoded	92.4 KiB	92.4 KiB	0 B (0.0%)	107.4 KiB	✅
Codex	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Codex	long turn finished, cold open	HTTP decoded	381.5 KiB	381.5 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, socket cold open	WebSocket decoded	381.6 KiB	381.6 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, work page	HTTP decoded	186.2 KiB	186.2 KiB	0 B (0.0%)	200.2 KiB	✅
Codex	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Codex	idle thread resume	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell snapshot	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Codex	fleet shell subscribe	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Codex	fleet shell resume check	rows	0	0	0	0	✅
Codex	stale shell reload	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Codex	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Codex	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	stale todo list cold open	HTTP decoded	72.4 KiB	72.4 KiB	0 B (0.0%)	83.0 KiB	✅
Codex	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Codex	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅
Claude	short turn resume	WebSocket decoded	10.6 KiB	10.6 KiB	0 B (0.0%)	12.7 KiB	✅
Claude	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn running, cold open	HTTP decoded	283.3 KiB	283.3 KiB	0 B (0.0%)	302.7 KiB	✅
Claude	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Claude	long turn running, history page	HTTP decoded	93.4 KiB	93.4 KiB	0 B (0.0%)	107.4 KiB	✅
Claude	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Claude	long turn finished, cold open	HTTP decoded	384.2 KiB	384.2 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, socket cold open	WebSocket decoded	384.3 KiB	384.3 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, work page	HTTP decoded	188.1 KiB	188.1 KiB	0 B (0.0%)	200.2 KiB	✅
Claude	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Claude	idle thread resume	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell snapshot	HTTP decoded	42.9 KiB	42.9 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Claude	fleet shell subscribe	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Claude	fleet shell resume check	rows	0	0	0	0	✅
Claude	stale shell reload	HTTP decoded	42.9 KiB	42.9 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Claude	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Claude	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	stale todo list cold open	HTTP decoded	73.4 KiB	73.4 KiB	0 B (0.0%)	83.0 KiB	✅
Claude	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Claude	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅

Baseline: 613d1ac · PR result: ab693a8 · Source CI: failure

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[coderabbitai]

No actionable comments were generated in the recent review.

Walkthrough

The OpenCode adapter now passes MCP availability to runtime instruction construction. Tests cover the disabled pull request linking option during MCP re-registration and an external-server session that writes runtime instructions and completes a turn.

Changes: Runtime instruction availability

Layer / File(s)	Summary
Set and verify instruction availability packages/provider-opencode/src/server/v2/adapter.ts, apps/server/src/orchestration-v2/Adapters/OpenCode2AdapterV2.test.ts	prepareTurn passes mcpAvailable as pullRequestLinking. Tests cover MCP re-registration and verify that an external-server session writes runtime instructions and completes a turn.

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix · Severity of issue fixed: Low

Suggested reviewers: yash-singh1

Merge Risk: Minimal · up to ab693

OpenCode sessions without T3 MCP tools will stop being told to call unavailable PR tools. The change is small and covered by tests, and no merge-blocking risk was identified.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title names the main change: sessions without T3 MCP no longer receive PR tool instructions. It follows the conventional commit form and matches the changeset, which passes pullRequestLinking: mcp…
Description check		The description follows the template well. It states the problem and the change, and it gives a scope section that links issue #35588 as the bug report. Verification is detailed, with test counts befo…
Linked Issues check		Issue #35588 is closed and completed. It supplies background only. No active directly linked target remains, so no linked-issue coding requirements apply. For reference, the whole-PR diff from merge b…
Out of Scope Changes check		The whole-PR diff from merge base 613d1ac to head ab693a8 touches two files. adapter.ts has the single call-site change that passes pullRequestLinking: mcpAvailable. OpenCode2AdapterV2.test.ts updat…

Full details: Title check

Explanation

The title names the main change: sessions without T3 MCP no longer receive PR tool instructions. It follows the conventional commit form and matches the changeset, which passes pullRequestLinking: mcpAvailable in the OpenCode v2 adapter.

Full details: Description check

Explanation

The description follows the template well. It states the problem and the change, and it gives a scope section that links issue #35588 as the bug report. Verification is detailed, with test counts before and after the fix, reproduction steps, and evidence. It is a little long and includes a non-template reproduction section, but all required sections are substantively complete.

Full details: Linked Issues check

Explanation

Issue #35588 is closed and completed. It supplies background only. No active directly linked target remains, so no linked-issue coding requirements apply. For reference, the whole-PR diff from merge base 613d1ac to head ab693a8 changes one call site in packages/provider-opencode/src/server/v2/adapter.ts. prepareTurn now passes pullRequestLinking: mcpAvailable to buildRuntimeInstructions, the same flag that gates the orchestration guidance. This matches the expected behavior in #35588. The external-server trust boundary is unchanged. The test file adds 'gives an external server's session no instructions for T3 tools it lacks' and updates the existing MCP re-registration expectation to pullRequestLinking: false. The PR description says the existing MCP-add-failed test also fails without the fix and passes with it. That test is not changed in this diff, so its coverage rests on the author's test report. This does not affect the verdict.

Full details: Out of Scope Changes check

Explanation

The whole-PR diff from merge base 613d1ac to head ab693a8 touches two files. adapter.ts has the single call-site change that passes pullRequestLinking: mcpAvailable. OpenCode2AdapterV2.test.ts updates one expectation to pullRequestLinking: false and adds one external-server test. Both changes connect directly to the reported instruction mismatch. The PR objectives mention an ACP adapter that already passes the option. No ACP file appears in the reviewed diff, so no out-of-scope code is introduced. The PR images and transcripts are evidence material and not repository changes.

Finishing Touches

Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Usage-based review receipt

• Mode: Continue automatically
• Reviewed files: 2
• Charged: $0.50
• View usage details

Note

This review was completed with usage-based billing: files reviewed beyond your plan's included limits are billed at $0.25/file. View usage-based billing.

Recent review info

Run configuration

• Source mode: Auto
• Repository: pingdotgg/t3code
• Branch: factory/issue-35588
• File: .coderabbit.config.ts
• Commit: ab693a8
• Layers: Repository TS config
• Review profile: CHILL
• Plan: Team
• Run ID: 65c995c8-c3ca-4db1-96da-4d07b9c61b43

Commits

Reviewing files that changed from the base of the PR and between 857fff9 and ab693a8.

Files selected for processing: 2

• apps/server/src/orchestration-v2/Adapters/OpenCode2AdapterV2.test.ts
• packages/provider-opencode/src/server/v2/adapter.ts

Limit details: You’ve used all 10 included reviews currently available.

[juliusmarminge]

The automated pipeline stopped on this pull request and will not merge it:

these checks fail on ab693a8, also when run again: Test Server 3, and the rework changed nothing:

No code change was needed. Item 1 (buildRuntimeInstructions called without pullRequestLinking: false) describes the code before this fix: at ab693a8, prepareTurn in packages/provider-opencode/src/server/v2/adapter.ts passes pullRequestLinking: mcpAvailable, so the <pull_request_linking> block is omitted whenever T3's MCP tools are unavailable. That covers both external servers and failed MCP registration, in the same way t3OrchestrationSystemPrompt is already gated. OpenCode2AdapterV2.test.ts asserts this: the MCP-unavailable case expects pullRequestLinking: false, and a new external-server case expects the same. Item 2 (Test Server 3 failed): CI's only failing test was apps/server/src/mcp/TaskCancelUsageLimit.integration.test.ts, which hit the 120s timeout. That test covers Claude delegated-task usage-limit cancellation and runs no OpenCode code. Locally it passed 4 out of 4 runs on this head, about 12s each. Recent main commits pass Test Server 3, so this was a CI timeout on that run, not caused by the change. A rerun of the job should clear it. The 'Check' job's only annotation was its own exit code 1.

It stays open for a person to finish or close.