[agento] [antig] refactor(agent-grok): switch from post-launch to inline prompt delivery#633
[agento] [antig] refactor(agent-grok): switch from post-launch to inline prompt delivery#633jleechan2015 wants to merge 25 commits into
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthroughBuild logic now appends the configured user prompt only when no explicit sessionId is supplied; tests were added/updated to verify inline system/user prompt placement on launch and suppression on resume/restore. Several small repository maintenance edits were made (imports, tests, lint, and pagination condition). ChangesGrok Inline Prompt Delivery
Repository maintenance and small fixes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 229a64b9f3
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
🧹 Nitpick comments (1)
packages/plugins/agent-grok/src/__tests__/index.test.ts (1)
214-237: 🏗️ Heavy liftSplit this test file into smaller suites to stay within the LOC guideline.
packages/plugins/agent-grok/src/__tests__/index.test.tsis now far above the ~300 LOC target; please split by concern (e.g., launch-command, activity-state, workspace-hooks) to keep maintenance and failures more localized.As per coding guidelines,
Keep files under approximately 300 lines of code (LOC) — split when it aids clarity.Also applies to: 509-509
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@packages/plugins/agent-grok/src/__tests__/index.test.ts` around lines 214 - 237, The test file is too large—split tests into focused suites (e.g., launch-command, activity-state, workspace-hooks): create new test files like launch-command.test.ts and move the tests that reference agent.getLaunchCommand and makeLaunchConfig into it; extract shared helpers (makeLaunchConfig, any shared setup/teardown) into a test-utils.ts module and import them into each suite; update imports in the moved tests to use the new test-utils and ensure each new file contains only the related describe/it blocks so no single test file exceeds ~300 LOC.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@packages/plugins/agent-grok/src/__tests__/index.test.ts`:
- Around line 214-237: The test file is too large—split tests into focused
suites (e.g., launch-command, activity-state, workspace-hooks): create new test
files like launch-command.test.ts and move the tests that reference
agent.getLaunchCommand and makeLaunchConfig into it; extract shared helpers
(makeLaunchConfig, any shared setup/teardown) into a test-utils.ts module and
import them into each suite; update imports in the moved tests to use the new
test-utils and ensure each new file contains only the related describe/it blocks
so no single test file exceeds ~300 LOC.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: ed664215-1751-4ef8-8341-9c0b660f65ac
📒 Files selected for processing (3)
packages/cli/__tests__/commands/spawn.test.tspackages/plugins/agent-grok/src/__tests__/index.test.tspackages/plugins/agent-grok/src/index.ts
|
@codex @coderabbitai @cursor @copilot [AI automation] Codex will implement the code updates while coderabbitai, cursor, and copilot focus on review support. Please make the following changes to this PR. Summary (Execution Flow):
PR Details:
Instructions: Tasks:
Automation Markers:
|
|
To use Codex here, create an environment for this repo. |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
# Fetch PR review comments
gh pr view 633 --repo jleechanorg/agent-orchestrator --json reviews,comments,reviewThreads 2>/dev/null | jq '.'
echo "---REVIEW_COMMENTS---"
gh api repos/jleechanorg/agent-orchestrator/pulls/633/comments 2>/dev/null | jq '.[] | {id, body, path, line, user: .user.login, html_url}'
echo "---ISSUE_COMMENTS---"
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null | jq '.[] | {id, body: .body[:300], user: .user.login, html_url}'Length of output: 2678 🏁 Script executed: #!/bin/bash
# Look at the Grok plugin index.ts on the PR branch
gh api repos/jleechanorg/agent-orchestrator/contents/packages/plugins/agent-grok/src/index.ts?ref=refactor/grok-inline-prompt-delivery 2>/dev/null \
| jq -r '.content' | base64 -dLength of output: 10951
Outstanding Review Comment[P2] Prompt not delivered when resuming Grok sessions on initial spawnComment: Assessment: Valid — should be fixed. Root cause: if (config.prompt && !restoreSessionId) {
parts.push("--", shellEscape(config.prompt));
}
const restoreSessionId = sessionId ?? getConfiguredGrokSessionId(config);When a project has Recommended fix for Codex: Distinguish between an initial spawn (where the prompt should always be delivered) and an explicit reconnect from // Before
if (config.prompt && !restoreSessionId) {
parts.push("--", shellEscape(config.prompt));
}
// After — suppress prompt only when explicitly reconnecting (sessionId arg is passed),
// always deliver on initial launch even if a configured session ID is being resumed
if (config.prompt && sessionId === undefined) {
parts.push("--", shellEscape(config.prompt));
}This preserves correct behaviour for both call sites:
A companion test case in No other outstanding review comments were found. All other threads are either from bots (auto-summary) or the automation trigger comment itself. |
jleechan2015
force-pushed
the
refactor/grok-inline-prompt-delivery
branch
from
229a64b to
c7f50c6
Compare
|
@coderabbitai all good? |
|
SKEPTIC_GATE_TRIGGER |
|
Tip For best results, initiate chat on the files or code changes.
[P2] Prompt silently dropped on initial launch when The current guard: if (config.prompt && !restoreSessionId) {
parts.push("--", shellEscape(config.prompt));
}…suppresses the prompt whenever Needed fix in // Only suppress the prompt when explicitly reconnecting (sessionId arg is passed by getRestoreCommand)
if (config.prompt && sessionId === undefined) {
parts.push("--", shellEscape(config.prompt));
}Plus a companion test in Once that is addressed, the PR should be good to merge. Full details are in my previous comment above. |
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@packages/plugins/agent-grok/src/__tests__/index.test.ts`:
- Around line 225-237: The test wrongly assumes setting
projectConfig.agentConfig.grokSessionId alone means an explicit reconnect;
update the assertions and test inputs so that only an explicit "resume" path
suppresses the prompt: in the "does not append prompt when resuming a session"
test (and the similar block at lines ~248-260) call agent.getLaunchCommand with
an explicit resume flag/option (the same param your production code uses to
trigger reconnect) when you expect prompt suppression, and for the initial-spawn
case keep grokSessionId set but do not set the resume flag so the generated
command (agent.getLaunchCommand and makeLaunchConfig usages) still contains the
appended prompt. Ensure you change the assertions accordingly (expect prompt
present for initial spawn with grokSessionId alone, and expect prompt absent
only when resume is explicitly set).
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: acab01d5-c376-4183-837e-cf670eb2d325
📒 Files selected for processing (2)
packages/plugins/agent-grok/src/__tests__/index.test.tspackages/plugins/agent-grok/src/index.ts
🚧 Files skipped from review as they are similar to previous changes (1)
- packages/plugins/agent-grok/src/index.ts
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@packages/plugins/agent-grok/src/index.ts`:
- Line 95: Change the loose equality check in the conditional that guards prompt
handling from "sessionId == null" to a strict undefined check "sessionId ===
undefined" to satisfy the eqeqeq lint while preserving behavior; locate the
conditional inside the buildGrokCommand / prompt-handling block in
packages/plugins/agent-grok/src/index.ts (the line starting with if
(config.prompt && sessionId == null)) and replace the comparison so callers that
omit the second arg (yielding undefined) still trigger the branch, and ensure no
other logic assumes null specifically.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 9e27d6a6-38b6-41dd-85da-5b811999855a
⛔ Files ignored due to path filters (1)
pnpm-lock.yamlis excluded by!**/pnpm-lock.yaml
📒 Files selected for processing (2)
packages/plugins/agent-grok/src/__tests__/index.test.tspackages/plugins/agent-grok/src/index.ts
🚧 Files skipped from review as they are similar to previous changes (1)
- packages/plugins/agent-grok/src/tests/index.test.ts
|
Addressed both CodeRabbit review comments:
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
# Check the current state of index.ts on the PR branch after the fixes
gh api repos/jleechanorg/agent-orchestrator/contents/packages/plugins/agent-grok/src/index.ts?ref=refactor/grok-inline-prompt-delivery 2>/dev/null \
| jq -r '.content' | base64 -d | grep -n -A5 -B5 "sessionId"Length of output: 2184
Confirmed in the current branch:
The logic is sound across all three scenarios from the original analysis:
This PR looks good to merge. 🐇 |
|
SKEPTIC_GATE_TRIGGER |
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
|
@codex @coderabbitai @cursor @copilot [AI automation] Codex will implement the code updates while coderabbitai, cursor, and copilot focus on review support. Please make the following changes to this PR. Summary (Execution Flow):
PR Details:
Instructions: Tasks:
Automation Markers:
|
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
# Check current state of the PR - review comments, issue comments, and CI checks
echo "=== PR REVIEW COMMENTS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/comments 2>/dev/null \
| jq '.[] | {id, body: .body[:400], user: .user.login, html_url, path, line, created_at}'
echo "=== PR ISSUE COMMENTS (last 5) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-5:] | .[] | {id, body: .body[:300], user: .user.login, html_url, created_at}'
echo "=== PR CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null | jq '.'
echo "=== RECENT COMMITS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[] | {sha: .sha[:8], message: .commit.message[:200], author: .commit.author.name, date: .commit.author.date}'Length of output: 9227 🏁 Script executed: #!/bin/bash
echo "=== CURRENT TEST FILE ==="
gh api repos/jleechanorg/agent-orchestrator/contents/packages/plugins/agent-grok/src/__tests__/index.test.ts?ref=refactor/grok-inline-prompt-delivery 2>/dev/null \
| jq -r '.content' | base64 -d
echo ""
echo "=== CURRENT INDEX.TS (relevant section) ==="
gh api repos/jleechanorg/agent-orchestrator/contents/packages/plugins/agent-grok/src/index.ts?ref=refactor/grok-inline-prompt-delivery 2>/dev/null \
| jq -r '.content' | base64 -d | grep -n -A8 -B8 "sessionId"
echo ""
echo "=== RESOLVE MY OWN COMMENTS ==="
# Get the IDs of my own review comments to resolve
gh api repos/jleechanorg/agent-orchestrator/pulls/633/comments 2>/dev/null \
| jq '.[] | select(.user.login == "coderabbitai[bot]") | {id, html_url, body: .body[:100]}'Length of output: 22755 🏁 Script executed: #!/bin/bash
# Resolve my own review comment threads via GraphQL
# First get the pull request node id and review thread IDs
echo "=== Getting PR node ID ==="
gh api graphql -f query='
{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
id
reviewThreads(first: 20) {
nodes {
id
isResolved
comments(first: 1) {
nodes {
databaseId
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
prId: .data.repository.pullRequest.id,
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
id,
isResolved,
commentId: .comments.nodes[0].databaseId,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:80]
}]
}'Length of output: 2187
Consolidated Review Status✅ All substantive issues addressed
CI / Green Gate
The code and tests are correct. The PR is ready for merge once the Green Gate re-evaluates against |
|
Codex Review: Didn't find any major issues. 🚀 ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
jleechan2015
force-pushed
the
refactor/grok-inline-prompt-delivery
branch
from
7d97092 to
86a6a0a
Compare
|
SKEPTIC_GATE_TRIGGER |
|
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes.
Based on my most recent analysis of the branch:
The PR is ready to merge. 🐇 |
|
Actionable comments posted: 0 |
|
Actionable comments posted: 0 |
… creation in direct-terminal-ws integration tests
|
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'
echo ""
echo "=== REVIEW THREADS RESOLUTION ==="
gh api graphql -f query='{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
reviewThreads(first: 30) {
nodes {
isResolved
comments(first: 1) {
nodes {
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
isResolved,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:150]
}]
}'
echo ""
echo "=== WebSocket test cleanup (direct-terminal-ws.test.ts) ==="
gh api "repos/jleechanorg/agent-orchestrator/contents/packages/web/server/__tests__/direct-terminal-ws.test.ts?ref=refactor/grok-inline-prompt-delivery" 2>/dev/null \
| jq -r '.content' | base64 -d | grep -n -A6 -B3 "closers\|ws\.close\|push" | head -80
echo ""
echo "=== LATEST ISSUE COMMENTS (last 3) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-3:] | .[] | {id, body: .body[:200], user: .user.login, created_at}'Length of output: 16143
Status on latest commit
|
| Item | Status |
|---|---|
| All review threads (9/9) | ✅ Resolved |
| WebSocket cleanup (Gate 7) | ✅ Fixed — closers.push(() => { socket.close(); }) added inside the Promise constructor before open fires, for both /ws and /ao-terminal-mux tests |
| CodeRabbit approval | 600b25a8 was auto-dismissed; re-approving now |
| CI (Lint, Test, Integration, Typecheck, etc.) | 🔄 Still running on 8c441f46 |
| Wholesome, Evidence, Dependency Review | ✅ SUCCESS |
All code concerns from the last Skeptic run are addressed. Re-approving to clear Gate 3 — once the in-progress CI checks resolve green, the Skeptic Gate should pass. 🐇
[approve]
… remote-tracking branch check priority
|
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'
echo ""
echo "=== REVIEW THREADS RESOLUTION ==="
gh api graphql -f query='{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
reviewThreads(first: 30) {
nodes {
isResolved
comments(first: 1) {
nodes {
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
isResolved,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:150]
}]
}'
echo ""
echo "=== LATEST SKEPTIC VERDICT (last 3 issue comments) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-3:] | .[] | {id, body: .body[:500], user: .user.login, created_at}'Length of output: 14042
Here's the current picture on latest commit
2 Unresolved Cursor Threads
Both concerns map directly to what Re-approving now. 🐇 [approve] |
…fy strict typescript-eslint rules
|
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS (last 5) ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'
echo ""
echo "=== REVIEW THREADS RESOLUTION ==="
gh api graphql -f query='{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
reviewThreads(first: 30) {
nodes {
isResolved
comments(first: 1) {
nodes {
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
isResolved,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:150]
}]
}'
echo ""
echo "=== LATEST SKEPTIC VERDICT (last 3 issue comments) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-3:] | .[] | {id, body: .body[:600], user: .user.login, created_at}'Length of output: 1737 🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS (last 5) ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'Length of output: 951 🏁 Script executed: #!/bin/bash
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state,reviews 2>/dev/null | jq '.'Length of output: 302
Could you let me know if the Skeptic Gate has re-run on |
|
SKEPTIC_GATE_TRIGGER |
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
Bugbot Autofix is ON. A cloud agent has been kicked off to fix the reported issue.
Reviewed by Cursor Bugbot for commit a0b5f0d. Configure here.
|
PR is green ✅ (6/6 criteria met — awaiting orchestrator merge) |
… from getRestoreCommand
|
@coderabbitai all good? |
|
SKEPTIC_GATE_TRIGGER |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS (last 5) ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'
echo ""
echo "=== REVIEW THREADS RESOLUTION ==="
gh api graphql -f query='{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
reviewThreads(first: 30) {
nodes {
isResolved
comments(first: 1) {
nodes {
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
isResolved,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:120]
}]
}'
echo ""
echo "=== LATEST SKEPTIC VERDICT (last 3 issue comments) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-3:] | .[] | {id, body: .body[:400], user: .user.login, created_at}'Length of output: 13630
Two new commits landed since my last approval (
These address cursor's last review thread ("Dead code with unsafe type casts in Current state on
|
| Item | Status |
|---|---|
| All review threads (12/12) | ✅ Resolved |
| CodeRabbit approval | |
| CI (Lint, Test, Integration, Typecheck, etc.) | 🔄 Still running on ea5de1cf |
| Static checks (Evidence, Secrets, Dependency Review, Actionlint) | ✅ SUCCESS |
| Skeptic Gate | ⏳ Not yet triggered on ea5de1cf |
Once CI completes and the Skeptic Gate re-evaluates against ea5de1cf, it should pass. Re-approving now. 🐇
[approve]
|
SKEPTIC_CRON_TRIGGER |
|
SKEPTIC_GATE_TRIGGER |
|
@coderabbitai all good? |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
echo "=== CURRENT PR STATE ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json headRefOid,mergeable,mergeStateStatus,state 2>/dev/null | jq '.'
echo ""
echo "=== CURRENT PR REVIEWS ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/reviews 2>/dev/null \
| jq '.[] | {id, user: .user.login, state, submitted_at, commit_id: .commit_id[:8]}'
echo ""
echo "=== RECENT COMMITS (last 5) ==="
gh api repos/jleechanorg/agent-orchestrator/pulls/633/commits 2>/dev/null \
| jq '.[-5:] | .[] | {sha: .sha[:8], message: .commit.message[:200], date: .commit.author.date}'
echo ""
echo "=== CI STATUS ==="
gh pr view 633 --repo jleechanorg/agent-orchestrator --json statusCheckRollup 2>/dev/null \
| jq '{checks: [.statusCheckRollup[] | {name, state, conclusion}]}'
echo ""
echo "=== REVIEW THREADS RESOLUTION ==="
gh api graphql -f query='{
repository(owner: "jleechanorg", name: "agent-orchestrator") {
pullRequest(number: 633) {
reviewThreads(first: 30) {
nodes {
isResolved
comments(first: 1) {
nodes {
author { login }
body
}
}
}
}
}
}
}' 2>/dev/null | jq '{
threads: [.data.repository.pullRequest.reviewThreads.nodes[] | {
isResolved,
author: .comments.nodes[0].author.login,
snippet: .comments.nodes[0].body[:120]
}]
}'
echo ""
echo "=== LATEST ISSUE COMMENTS (last 3) ==="
gh api repos/jleechanorg/agent-orchestrator/issues/633/comments 2>/dev/null \
| jq '.[-3:] | .[] | {id, body: .body[:400], user: .user.login, created_at}'Length of output: 14465
A new commit landed since my last approval:
This auto-dismissed my approval on Current state on
|
| Item | Status |
|---|---|
| All review threads (12/12) | ✅ Resolved |
| CodeRabbit approval | |
| Static checks (Evidence, Wholesome, Secrets, Dependency Review, Actionlint, NPM Audit, etc.) | ✅ SUCCESS |
| Lint / Typecheck / Test / Integration Tests / Diff Coverage / Cursor Bugbot | 🔄 Still running on c1b4ad4d |
Once CI completes green, the Skeptic Gate should re-evaluate and pass. 🐇
[approve]
Green Gate — Deterministic 6-Green Check
Result: PASS — all gates green VERDICT: PASS |
|
SKEPTIC_GATE_TRIGGER |
2 participants
Background
PR #633: [agento] [antig] refactor(agent-grok): switch from post-launch to inline prompt delivery — refactors prompt delivery in the agent-grok plugin to occur at spawn time via command line arguments rather than post-launch terminal input stream. It also hardens base branch resolution in packages/core, validates limit query parameters against
NaNin web sessions API, and ensures proper cleanup of WebSocket connections in integration tests.Goals
resolveBaseBranchto guarantee wholesome tests are executed against up-to-date refs in CI, and prevent GITHUB_BASE_REF environment pollution.limit=NaN) is supplied.Tenets
High-Level Description
packages/plugins/agent-grok:
buildGrokCommandappends the task prompt inline after--using proper shell escaping, but ONLY when first launching the session (sessionId == null).getRestoreCommandomits/suppresses the prompt since the session already exists.getRestoreCommand, satisfying stricttypescript-eslintrules.packages/core:
resolveBaseBranchvalidates remote-tracking branches (origin/${validated}) first before checking local references.git fetchspecifically to fetch the remote-tracking ref if it is missing locally.wholesome.test.tsto prevent process.env pollution.packages/web:
/api/sessionsAPI checks query limits using typesafetypeof limit === "number" && !isNaN(limit).closersarray and close them inafterEachhooks to prevent dangling processes.Testing
Evidence
CI Run Proofs (HEAD Commit: c1b4ad4da425dc8e672f8cf1f61fe6ed25d3270b)
Real Test Execution Logs
1. Grok Inline Prompt & Reconnect Verification
2. Wholesome Core Base Branch Verification
3. Web Server & API Compatibility Verification
$ pnpm -C packages/web test compatibility vitest run compatibility RUN packages/web ✓ server/__tests__/server-compatibility.test.ts (15 tests) 11ms Test Files 1 passed (1) Tests 15 passed (15)4. WebSocket Server & Routing Integration Verification
$ pnpm -C packages/web test direct-terminal-ws vitest run direct-terminal-ws RUN packages/web ✓ server/__tests__/direct-terminal-ws.test.ts (4 tests) 227ms Test Files 1 passed (1) Tests 4 passed (4)Design doc: pr-633.md
Bead: N/A
Note
Medium Risk
Changes Grok launch/reconnect command shaping (agent spawn path) and wholesome git base resolution used for branch diff gates; sessions API pagination behavior changes only for invalid numeric limits.
Overview
agent-grok now delivers the task prompt inline on the launch command (positional arg after
--) instead of post-launch terminal injection. Prompt inclusion is gated onsessionId == null, so initial spawns still get the prompt whengrokSessionIdis configured, whilegetRestoreCommandreconnects omit it to avoid duplicate prompts.Wholesome / CI diff base:
resolveBaseBranchprefersorigin/*, can shallow-fetch a missingGITHUB_BASE_REF, and resolves via a lazy cached getter so Vitest discovery does not run git at import time. Tests restoreGITHUB_BASE_REFand add a resilience case.Web:
GET /api/sessionstreats invalidlimit(includingNaN) as unpaginated viatypeof limit === "number" && !isNaN(limit). Direct-terminal WebSocket tests register afterEach closers and cover/wsand/ao-terminal-mux; compatibility tests assertnoServer: trueon the WSS.Notifier tests document retry behavior: 400/422 do not retry; 429 does. A PR design doc (
pr-633.md) captures rationale;spawn-queuehas only whitespace.Reviewed by Cursor Bugbot for commit c1b4ad4. Bugbot is set up for automated code reviews on this repo. Configure here.