[juliusmarminge]

Problem

ACP Registry command discovery stops after the 128th advertised command. normalizeAcpRegistryCommands counts accepted entries and breaks at a hard-coded MAX_COMMANDS = 128, so a Devin setup with a large skill catalog silently loses every later skill from the / picker. Both the background probe and live available_commands_update notifications go through it. On main, 140 advertised skills come back as 128 and the 140th (poteto-mode) is missing. $-prefixed skills count against the same cap.

Change

Remove the count cap from normalizeAcpRegistryCommands in packages/provider-acp-registry/src/server/probe.ts. These stay as they were: the per-entry name, description and hint length bounds, case-insensitive de-duplication, and $ skill routing. ACP Registry now keeps every valid, distinct advertised command, as the Grok and generic ACP adapters already do. The ServerProvider.slashCommands/skills contracts have no item limit.

Scope and approval

Fixes #33757

Verification

• vp test run src/provider/AcpRegistryProbe.test.ts (apps/server): the existing "bounds and de-duplicates advertised commands" case is now "keeps every distinct advertised command and de-duplicates the rest". It advertises 143 entries: create_plan, a case-duplicate CREATE_PLAN, 140 generated commands and a $late-skill at the end. It expects 141 slash commands plus the skill.
  • On main it fails with expected [ …(128) ] to have a length of 141 but got 128.
  • With the fix all 11 tests in the file pass.
• vp run check:changed on the merged head 582b3dc: format, lint, knip, typecheck (10 packages) and tests (2 files, 13 tests) pass (transcript).
• Runtime transcript: a synthetic stdio ACP agent advertising 143 entries, run through the real probeAcpRegistryConfiguration.
  • Main 633438d keeps 128 (128 !== 141): runtime-main-633438d54f.log
  • Head 0972902 keeps 141 slash commands plus late-skill, with the duplicate removed: runtime-head-09729029ba.log
  • Drivers (synthetic, not committed): runtime.mjs, catalog-agent.mjs
• Normalizer driver, three variants:
  • The reporter's case: 140 unprefixed skills, target last.
  • 128 commands followed by one $ skill.
  • 129 commands.
  • Main gives 128/absent, 0 skills and 128 (log). Head gives 140/present, 1 skill and 129 (log). Driver: normalizer-repro.ts.
• Cost: the loop is a single linear pass. 100k distinct commands normalize in about 127 ms, and 3,000 malformed entries in about 13 ms (results, driver).
• No visible UI change, so no captures.
• Not checked: a live Devin session; the reporter's catalog was not available.

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

[macroscopeapp]

Approvability

Verdict: Approved at 582b3dc

Macroscope's review found this PR approvable — ACP command discovery now retains valid commands and skills beyond the former 128-entry limit. Existing validation and deduplication remain, and the change affects advertised metadata rather than automatically executing additional commands.

^{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	✅

Baseline: e4d86ab · PR result: 582b3dc · Source CI: success

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 ACP Registry command normalizer no longer stops after 128 accepted commands. The updated test checks that 141 slash commands are retained and that a later $late-skill entry appears as a skill.

Changes: ACP Registry command discovery

Layer / File(s)	Summary
Normalize the full command list packages/provider-acp-registry/src/server/probe.ts, apps/server/src/provider/AcpRegistryProbe.test.ts	The normalizer processes all input commands without the former 128-command limit. The test expects 141 slash commands and verifies that $late-skill is classified as late-skill.

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

Change: Bug fix · Severity of issue fixed: Medium

Suggested reviewers: t3dotgg

Merge Risk: Minimal · up to 582b3

Commands beyond the former limit are retained, with no merge-blocking issue identified.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title clearly describes the main change: commands beyond the former 128-command limit are retained.
Linked Issues check		Issue #33757 requires all valid advertised commands to remain discoverable. The PR removes the 128-command early exit from normalizeAcpRegistryCommands and retains the existing filtering and de-dupl…
Out of Scope Changes check		The PR changes only the ACP Registry command normalizer and its focused test. Both changes directly support issue #33757. The PR retains no command-count bound, so the issue's alternative requirement …
Description check		The description covers the problem, change, and verification in detail. It links issue #33757, but does not include explicit maintainer approval or explain why the fix qualifies for the small, obvious…

Full details: Linked Issues check

Explanation

Issue #33757 requires all valid advertised commands to remain discoverable. The PR removes the 128-command early exit from normalizeAcpRegistryCommands and retains the existing filtering and de-duplication. The updated test checks 141 slash commands and a later $late-skill, which remains routed to skills.

Full details: Out of Scope Changes check

Explanation

The PR changes only the ACP Registry command normalizer and its focused test. Both changes directly support issue #33757. The PR retains no command-count bound, so the issue's alternative requirement to report truncation does not apply.

Full details: Description check

Explanation

The description covers the problem, change, and verification in detail. It links issue #33757, but does not include explicit maintainer approval or explain why the fix qualifies for the small, obvious bug exemption.

Finishing Touches

Generate docstrings

• Commit to this branch
• Create a new PR

Generate unit tests (beta)

• Commit to this branch
• Create a new PR

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

Recent review info

Run configuration

• Source mode: Auto
• Repository: pingdotgg/t3code
• Branch: fix/issue-33757
• File: .coderabbit.config.ts
• Commit: 582b3dc
• Layers: Repository TS config
• Review profile: CHILL
• Plan: Team
• Run ID: 9c9cdf1a-0889-410e-b3bd-8451028927ac

Commits

Reviewing files that changed from the base of the PR and between e4d86ab and 582b3dc.

Files selected for processing: 2

• apps/server/src/provider/AcpRegistryProbe.test.ts
• packages/provider-acp-registry/src/server/probe.ts

Files with no reviewable changes: 1

• packages/provider-acp-registry/src/server/probe.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 0 remain after this review.