Two blockers from the maintainer's Oracle review on PR #4121:
Blocker 1 — load_skills=null should still throw, omitted should default
The previous PR collapsed both `loadSkills === undefined` and
`loadSkills === null` into a silent default of `[]`. The closing
rationale of PR #1663 (which reverted PR #1493) and the maintainer's
review both call out the importance of preserving the distinct
"omitted -> default, explicit invalid -> throw" contract. `null`
strongly signals "I tried to pass something and it was wrong";
silently coercing it hides bugs upstream.
Restored the split: `undefined` -> default `[]` + log,
`null` -> throw with the historical error string.
Blocker 2 — task_id continuation test rewritten, not deleted
The original PR removed the `task_id without run_in_background ->
throws` test entirely. The behavior IS preserved (default false ->
`isExplicitSyncRun` true -> `executeSyncContinuation`), but with the
test gone the new contract was unprotected.
Added a regression test that asserts the new contract: when
`task_id` is present and `run_in_background` is omitted,
`tool.execute` must route through sync continuation without throwing
the legacy required-parameter error. Mocks include `session.abort`
because the sync poller calls it during shutdown.
Also flipped the existing `load_skills=null` regression test from
"normalizes to []" back to "throws with the legacy error string" to
match the restored contract.
Tests:
- bun test src/tools/delegate-task/tools.test.ts -> 132/132 pass
- bun test src/tools/delegate-task/ -> 406/406 pass
- bun run typecheck -> clean
Sisyphus and other delegators occasionally invoke the task() tool without
an explicit run_in_background or load_skills argument. The runtime
validators in tool-argument-preparation.ts threw a hard Error in that
case, which short-circuited tool.execute() entirely. Because OpenCode's
tool.execute.after hook only runs on returned results, the
delegate-task-retry hook never had a chance to attach corrective
guidance — so the model saw a raw failure and either burned several
retries or fell back to a synchronous Explore call, silently losing
parallel execution.
Behavior change:
- run_in_background omitted -> defaults to false (sync delegation), with
a log entry for observability.
- load_skills omitted or null -> normalized to [] with a log entry on
the explicit-null path.
- The Zod schema entries are now .optional() and their .describe()
strings declare the defaults honestly; the markdown tool description
was updated to match (no more 'REQUIRED' lie).
The orthogonal validation 'Must provide either category or
subagent_type.' is unchanged and still surfaces as a returned error.
Tests:
- The five throw-on-missing tests in tools.test.ts are rewritten to
assert the new default-and-proceed contract.
- The 'no category, no subagent_type' test now asserts the
missing-target error remains intact.
Refs the workaround the reporter validated in the original issue body;
matches the design from PR #2375 which was previously reverted by
566031f4.
Include Vertex Anthropic in the assistant-tail prefill guard after Cubic flagged the strict provider check. Keep the non-Anthropic opencode negative control covered by tests.
Plan: plans/fix-anthropic-assistant-prefill-tail.md
Tighten the assistant-tail repair after review so only the Anthropic provider is treated as prefill-rejecting, and so assistant-tail model metadata is evaluated independently from the last user model.
Add regression coverage for an allowed user model followed by a rejecting Anthropic assistant tail, plus a non-Anthropic provider carrying a Claude-looking model id.
Plan: plans/fix-anthropic-assistant-prefill-tail.md
Claude Opus 4.7 rejects requests whose final message is an assistant turn, because Anthropic treats that as unsupported assistant prefill. Add a transform guard that appends the existing synthetic user recovery turn only for Anthropic prefill-rejecting model IDs, while preserving assistant-prefill behavior for supported and unknown models.
Tests cover Opus 4.7, Opus 4.6 assistant metadata fallback, direct Sonnet 4.6 model fields, allowed OpenAI/Sonnet 4.5 tails, missing metadata, and compaction continuation compatibility.
Plan: plans/fix-anthropic-assistant-prefill-tail.md
Discord support and other guidance still tell users 'bunx oh-my-opencode setup', but the CLI only registers 'install'. Running the recommended command produces:
error: unknown command 'setup'
which dead-ends users who are trying to refresh missing skills (hyperplan, etc.) after upgrading. Commander.js supports first-class aliases, so declaring '.alias("setup")' on the install command makes both names route to the same action without any code duplication. Help output now reads 'Usage: oh-my-opencode install|setup [options]', which is the right discovery hint for both newcomers and users following older docs.
Regression test parses cli-program.ts and asserts the install command block contains '.alias("setup")' before its '.action(' so a future refactor of the CLI cannot silently drop the alias.
ensureBaseDirs unconditionally called chmod(baseDir, 0o700) on every startup
and on every team_create. On filesystems where the OS rejects chmod for the
directory (network mounts, SIP-protected locations, non-owner cases on macOS
shared by multiple GUI users), the call raises EPERM and the entire team-mode
init aborts:
[team-mode] init failed: EPERM: operation not permitted, chmod '/Users/<u>/.omo'
Wrap chmod through a small safeChmod helper that converts EPERM, ENOTSUP, and
EINVAL into a single warning log and continues. mkdir already creates new
directories with mode 0o700, and the existing post-creation stat-guard remains
in place for the case where the directory pre-exists with a different mode and
chmod is permitted, so the security envelope on supported filesystems is
unchanged. All other error codes (ENOENT, EACCES, etc.) still propagate.
Regression test mocks node:fs/promises.chmod to throw EPERM and asserts that
ensureBaseDirs completes successfully and emits exactly the documented warning.
The multimodal-looker prompt described what to do but never told the model which tools are available. Smaller VL models (e.g. Qwen3-VL-8B) would try to call non-existent tools and enter an infinite loop emitting:
Model tried to call unavailable tool 'invalid'. Available tools: call_omo_agent, read.
Add a single sentence at the top of the prompt that explicitly enumerates the only allowed tools ('read' and 'call_omo_agent') and forbids calling any other tool. This matches the runtime allowlist enforced by createAgentToolAllowlist(["read"]).
Regression test asserts the prompt contains the available-tools enumeration so future prompt rewrites don't regress.
Require the latest assistant tool-wait turn itself to be stale before a shouldReply parent wake can bypass tool-call deferral. This prevents an all-complete background wake from forking a second parent prompt loop when OpenCode has repaired the tail to a synthetic user message.
Tests:
- bun test src/features/background-agent/parent-wake-user-message-race.test.ts src/features/background-agent/task-completion-cleanup.test.ts src/hooks/shared/prompt-async-gate.test.ts src/shared/prompt-async-route-audit.test.ts --bail
- bun run typecheck
- bun test
Keep the built-in lsp MCP registered even when the submodule CLI artifact is missing.
The fallback command initializes the lsp-tools-mcp submodule, prefers the source CLI without dirtying the checkout with dist output, and keeps npm build as a last resort when Bun cannot run the source entrypoint.
Plan: plans/fix-lsp-mcp-missing-cli.md
The bundled createTodoDescriptionOverrideHook returns { 'tool.definition': fn }, but plugin-interface.ts never exposes 'tool.definition' as an OpenCode hook handler. Result: the hook is constructed by createToolGuardHooks (line 132-134 of src/plugin/hooks/create-tool-guard-hooks.ts) but the function is never invoked, so todowrite keeps using OpenCode's core 7 KB description instead of the 1.4 KB TODOWRITE_DESCRIPTION. User-defined plugins under ~/.config/opencode/plugin/*.js use the same hook contract and work fine, confirming the contract itself is functional in opencode 1.14.28+.
Fix: add src/plugin/tool-definition.ts (createToolDefinitionHandler) that forwards the OpenCode 'tool.definition' input/output pair into hooks.todoDescriptionOverride. Wire it into plugin-interface.ts alongside tool.execute.before/after.
Regression coverage: src/plugin/tool-definition.test.ts covers (a) todowrite override applied, (b) other tools left untouched, (c) null hook is a no-op.
After the 4.2.0 unified-dispatch refactor (a42f894f / df198d8b / fee515c5 / 989ab717 / dd3fecaf / 1bbe065c / 12bd6580), at least one caller in the new prompt-async-gate path forwards a FallbackModelObject (or some other non-string shape) into parsers that statically claim 'model: string'. The downstream .trim() call then throws 'model.trim is not a function', which rejects the session.processor promise and surfaces as 'Aborted process' + UI 'interrupted'. The issue (#4145) reports this aborts 90% of subagent dispatches across every provider on 4.2.0 + opencode 1.15.4.
This patch adds a 'typeof x !== "string"' runtime guard at the four parser entrypoints called from the dispatch path:
- src/shared/fallback-chain-from-models.ts :: parseVariantFromModel, parseFallbackModelEntry
- src/tools/delegate-task/model-string-parser.ts :: parseVariantFromModelID, parseModelString
- src/shared/model-string-parser.ts (duplicate file with same API) :: parseVariantFromModelID, parseModelString
- src/features/claude-code-agent-loader/claude-model-mapper.ts :: mapClaudeModelString
Each parser now returns undefined / { modelID: "" } for non-string input instead of throwing. This unblocks subagent dispatch and leaves the underlying caller bug for a follow-up.
Regression coverage: three new tests in src/shared/fallback-chain-from-models.test.ts pin the non-string behavior (object, null/undefined, number). Existing 38 tests still pass. Total: 41/41 green, typecheck clean.
When OpenCode reports the original assistant error after a fallback retry has already been accepted, keep waiting for the pending fallback model instead of clearing the awaiting flag.
This prevents a duplicate stale session.error from advancing the fallback chain and dispatching a second assistant retry prompt.
Claude Agent