chore(acp): bump dirac and nova registry pins to probed versions - #750
Merged
Conversation
kaizhou-lab
added a commit
that referenced
this pull request
Aug 3, 2026
) ## Summary Scheduled ACP Registry version sync. Two npx pins drifted since #750, each backed by a fresh serial ACP probe of the exact pinned version. Lock-only change — two lines; no metadata, migration, or entrypoint edits, and no lock-derived test assertion embeds either version. | backend | package | old → new | initialize | session/new | |---|---|---|---|---| | dimcode | `dimcode` | 0.3.1 → **0.3.2** | ok (agentInfo dimcode "DimCode" 0.3.2) | auth required (`-32000`, "Provider credentials are required") | | grok | `@xai-official/grok` | 0.2.118 → **0.2.119** | ok (protocolVersion 1) | auth required (`-32000`, "no auth method id provided") | Both meet the release-lock criterion: `initialize` succeeds and `session/new` returns a clearly classified authentication requirement. Probes ran serially with no inherited HOME or credentials. The other 9 Registry-pinned packages (autohand, codebuddy, deepagents, dirac, glm-acp-agent, kilo, nova, pi, sigit) match the snapshot exactly — no drift. Package names and entrypoint args are unchanged for all 11. Drifted but not upgraded: none. The `mimo-code` and `omp` entries are non-Registry builtins (no `registry_json_id`) and are correctly excluded from drift reconciliation. ## Registry snapshot - Audit pinned to release tag [`v2026.08.03-e5290cd`](https://cdn.agentclientprotocol.com/registry/v1/v2026.08.03-e5290cd/registry.json) of `agentclientprotocol/registry` (newest release at audit time), fetched via the versioned CDN path for reproducibility. - No newly listed and no delisted agents versus the previous baseline (38 ids). ## Validation - `cargo test -p aionui-runtime -p aionui-ai-agent` — pass (882 lib + integration tests) - Full `just push` gate (migration check, lint, fmt, `cargo nextest run --workspace`) — pass - Gate run with the host-injected `AIONUI_LOG_DIR` unset (pre-existing environment sensitivity, see #695) - Gate started only after host load fell below the pre-flight threshold; the targeted test run had pushed the 12-core host to load 79, which is the condition that produced spurious subprocess-spawn timeouts in #750's history. ## Logging No logging changes: lock-only version bumps; existing startup/session error paths already identify a failing agent by backend. Co-authored-by: zk <>
kaizhou-lab
added a commit
that referenced
this pull request
Aug 5, 2026
## Summary Scheduled ACP Registry version sync. Two npx pins drifted since #765, each backed by a fresh serial ACP probe of the exact pinned version. Lock-only change — two lines; no metadata, migration, or entrypoint edits, and no lock-derived test assertion embeds either version. | backend | package | old → new | initialize | session/new | |---|---|---|---|---| | grok | `@xai-official/grok` | 0.2.119 → **0.2.120** | ok (protocolVersion 1) | auth required (`-32000`, "no auth method id provided") | | kilo | `@kilocode/cli` | 7.4.17 → **7.4.19** | ok (agentInfo Kilo 7.4.19) | success | Both meet the release-lock criterion: `initialize` succeeds and `session/new` either succeeds or returns a clearly classified authentication requirement. Probes ran serially with no inherited HOME or credentials. The other 9 Registry-pinned packages (autohand, codebuddy, deepagents, dimcode, dirac, glm-acp-agent, nova, pi, sigit) match the snapshot exactly — no drift. Package names and entrypoint args are unchanged for all 11. Drifted but not upgraded: none. The `mimo-code` and `omp` entries are non-Registry builtins (no `registry_json_id`) and are correctly excluded from drift reconciliation. ## Registry snapshot - Audit pinned to release tag [`v2026.08.04-c23c1ec`](https://cdn.agentclientprotocol.com/registry/v1/v2026.08.04-c23c1ec/registry.json) of `agentclientprotocol/registry` (newest release at audit time), fetched via the versioned CDN path for reproducibility. - No newly listed and no delisted agents versus the previous baseline (38 ids). ## Validation - `cargo test -p aionui-runtime -p aionui-ai-agent` — pass (902 lib + integration tests) - Full `just push` gate (migration check, lint, fmt, `cargo nextest run --workspace`) — pass - Gate run with the host-injected `AIONUI_LOG_DIR` unset (pre-existing environment sensitivity, see #695) - Gate started only after host load fell below the pre-flight threshold; the targeted test run had pushed the 12-core host to load 69, the condition that produced spurious subprocess-spawn timeouts in #750's history. ## Logging No logging changes: lock-only version bumps; existing startup/session error paths already identify a failing agent by backend. Co-authored-by: zk <>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Scheduled ACP Registry version sync. Two npx pins drifted since #749, each backed by a fresh serial ACP probe of the exact pinned version. Lock-only change — two lines; no metadata, migration, or entrypoint edits, and no lock-derived test assertion embeds either version.
dirac-cli@compass-ai/nova-32000, "Click Nova Setup to configure your API keys", advertising akore-terminal-authterminal method)Both meet the release-lock criterion:
initializesucceeds andsession/neweither succeeds or returns a clearly classified authentication/configuration requirement. Probes ran serially with no inherited HOME or credentials.The other 9 Registry-pinned packages (autohand, codebuddy, deepagents, dimcode, glm-acp-agent, grok, kilo, pi, sigit) match the snapshot exactly — no drift. Package names and entrypoint args are unchanged for all 11. Drifted but not upgraded: none. The
mimo-codeandompentries are non-Registry builtins (noregistry_json_id) and are correctly excluded from drift reconciliation.Carried forward from #739: nova still self-reports
agentInfo.name = "kore-cli",version = "1.0.0", matching neither the product name nor the package version. Unchanged behaviour, not a blocker — the Registry/public-page identity and package are stable andinitializesucceeded.Registry snapshot
v2026.08.02-27d86f5ofagentclientprotocol/registry(newest release at audit time), fetched via the versioned CDN path for reproducibility.Validation
cargo test -p aionui-runtime -p aionui-ai-agent— pass (831 lib + integration tests)just pushgate (migration check, lint, fmt,cargo nextest run --workspace) — passAIONUI_LOG_DIRunset (pre-existing environment sensitivity, see chore(acp): bump registry npx lock to probed versions #695)Note on validation history: four earlier gate attempts on this same commit failed, each on a different subprocess-spawning test (
aionui-mcpPATH resolution,aionui-extensionlifecycle hooks ×2,aionui-appwebsocket e2e) and each landing exactly on that test's own timeout (30s / 120s / 26.8s). Those were host-contention artifacts, not defects: macOSsyspolicyd,XProtectandStorageManagementheld the 12-core host near load average 300 for over an hour, stalling everyexec(), and the same tests pass isolated in 0.44s and 1.9s. The gate passed unchanged once load fell to ~5; no source change was needed.Logging
No logging changes: lock-only version bumps; existing startup/session error paths already identify a failing agent by backend.