fix(agent): hide code mode wait progress - #100164

Best possible solution:

Land the adapter-preservation fix after normal protected-label maintainer review, keeping suppression driven by the existing metadata flag instead of tool-name special cases.

Do we have a high-confidence way to reproduce the issue?

Yes. Source inspection shows Code Mode wait tools set hideFromChannelProgress, while current main drops that flag in the AgentTool/ToolDefinition adapter paths before channel-progress handling can honor it.

Is this the best way to solve the issue?

Yes. Preserving the existing boolean through the adapter boundaries is the narrow maintainable fix and avoids fragile name-based hiding for unrelated tools named wait.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against 420b3cdd2936.

Label changes:

• add P2: This is a normal-priority user-visible Discord progress bug with a narrow agent-runtime fix and limited blast radius.
• add proof: sufficient: Contributor real behavior proof is sufficient. The PR body reports after-fix live gateway proof showing repeated internal waits no longer produced Discord Wait rows, alongside focused regression and full-check validation.
• add rating: 🦞 diamond lobster: Overall readiness is 🦞 diamond lobster; proof is 🦞 diamond lobster and patch quality is 🦞 diamond lobster.
• add status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR body reports after-fix live gateway proof showing repeated internal waits no longer produced Discord Wait rows, alongside focused regression and full-check validation.

Label justifications:

• P2: This is a normal-priority user-visible Discord progress bug with a narrow agent-runtime fix and limited blast radius.
• rating: 🦞 diamond lobster: Overall readiness is 🦞 diamond lobster; proof is 🦞 diamond lobster and patch quality is 🦞 diamond lobster.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR body reports after-fix live gateway proof showing repeated internal waits no longer produced Discord Wait rows, alongside focused regression and full-check validation.
• proof: sufficient: Contributor real behavior proof is sufficient. The PR body reports after-fix live gateway proof showing repeated internal waits no longer produced Discord Wait rows, alongside focused regression and full-check validation.

PR surface:

Source +5, Tests +48. Total +53 across 5 files.

What I checked:

• Current Code Mode marker: Current main marks the internal Code Mode wait control with hideFromChannelProgress: true, so the intended behavior is metadata-based suppression rather than hiding any tool named wait. (src/agents/code-mode.ts:1172, 420b3cdd2936)
• Current channel lifecycle contract: Current main's agent-core AgentTool type includes hideFromChannelProgress, and the loop uses it to tag tool lifecycle events for channel-progress consumers. (packages/agent-core/src/types.ts:465, 420b3cdd2936)
• Adapter gap on current main: toToolDefinitions currently copies name, label, description, parameters, prepareArguments, and executionMode but not hideFromChannelProgress, which explains why the marker can be lost after SDK tool conversion. (src/agents/agent-tool-definition-adapter.ts:367, 420b3cdd2936)
• Wrapper gap on current main: The ToolDefinition wrapper currently omits hideFromChannelProgress in both ToolDefinition-to-AgentTool and AgentTool-to-ToolDefinition directions, matching the PR's stated adapter-boundary problem. (src/agents/sessions/tools/tool-definition-wrapper.ts:19, 420b3cdd2936)
• PR diff scope: The PR adds only metadata propagation at the three missing adapter points, adds the optional ToolDefinition field, and adds focused regression tests that preserve hidden internal waits while leaving unmarked wait-named tools visible. (src/agents/agent-tool-definition-adapter.ts:367, 4d65163017eb)
• Related merged provenance: Merged pull request fix(discord): hide internal Code Mode wait progress #99893 introduced the initial Code Mode wait progress suppression on July 4, 2026; this PR is a narrower follow-up for the adapter path that remained uncovered. (packages/agent-core/src/agent-loop.ts:618, 3d404478b802)

Likely related people:

• steipete: Authored and merged the initial Code Mode wait progress suppression in fix(discord): hide internal Code Mode wait progress #99893 and has recent agent-core/code-mode history on the same behavior. (role: recent feature owner; confidence: high; commits: 3d404478b802, 5ea80c8d80fd; files: src/agents/code-mode.ts, packages/agent-core/src/agent-loop.ts, src/auto-reply/reply/agent-runner-execution.ts)
• vincentkoc: Recent commits touched the agent tool adapter/code-mode surfaces and adjacent agent runtime behavior, making this a plausible routing candidate for adapter-boundary review. (role: recent adjacent area contributor; confidence: medium; commits: b832dd27e1cc, e7e686db2d00, 2e44610ba237; files: src/agents/agent-tool-definition-adapter.ts, src/agents/code-mode.ts)

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.