Fix usage summary, replace parallel plugin with sequential review - #1507
Merged
Conversation
The execution file is a list of messages, not a single object. Scan the array for the entry containing modelUsage. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
|
Important Installation incomplete: to start using Gemini Code Assist, please ask the organization owner(s) to visit the Gemini Code Assist Admin Console and sign the Terms of Services. |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #1507 +/- ##
==========================================
+ Coverage 92.21% 93.32% +1.10%
==========================================
Files 76 76
Lines 8233 8162 -71
==========================================
+ Hits 7592 7617 +25
+ Misses 641 545 -96
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Contributor
Code reviewNo issues found. Checked for bugs and CLAUDE.md compliance. 🤖 Generated with Claude Code - If this code review was useful, please react with 👍. Otherwise, react with 👎. |
Fork of Anthropic's code-review plugin but sequential instead of parallel subagents. Same quality checks (CLAUDE.md compliance, bug detection, confidence scoring, false positive filtering) but reuses cached context instead of duplicating across 5+ parallel agents. Previous run: $2.77 for a small PR. Also adds aiobotocore-specific async pattern checking. Removes plugin_marketplaces and plugins from workflow. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
thehesiod
changed the title
thehesiod
added a commit
that referenced
this pull request
Apr 19, 2026
The Claude Code GitHub Action does not auto-discover loose
.claude/commands/*.md files, so every /review-pr invocation on this
branch failed with "Commands are in the form \`/command [args]\`" —
every Claude CI run returning in 0s without actually reviewing the
PR. This packages the commands as a Claude Code plugin the action
can register explicitly.
Structure (mirrors Anthropic's own plugins/code-review/ layout):
.claude-plugin/marketplace.json # marketplace declaration
plugins/aiobotocore-bot/
.claude-plugin/plugin.json # plugin manifest
README.md
commands/
review-pr.md # moved from .claude/commands/
analyze-pr-feedback.md # moved from .claude/commands/
Workflow change in .github/workflows/claude.yml:
plugin_marketplaces: ./.
plugins: aiobotocore-bot@aiobotocore
Uses a local filesystem path (repo root) rather than a git URL, made
possible by anthropics/claude-code-action#761. This matters because
the action installs from the checked-out working tree, so PR branches
test their own plugin edits in their own CI run — same property as
the envsubst-from-worktree behavior for the prompt templates.
Commands become namespaced (per Claude Code plugin convention):
/review-pr → /aiobotocore-bot:review-pr
/analyze-pr-feedback → /aiobotocore-bot:analyze-pr-feedback
Also adopts structural improvements from the official code-review
plugin: allowed-tools: frontmatter restricting each command's tool
surface, agent-assumptions statement. We keep our sequential review
loop (cost constraint from #1507) and our aiobotocore-specific async
pattern check rather than switching to the official parallel flow.
Updates .github/claude-review-prompt.md and docs/ai-workflows.md to
reference the new paths and namespaced commands. No change to
botocore-sync.yml — it does not use either slash command.
9 tasks
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
Two changes:
Fix usage-summary.py
The execution_file is a JSON array, not a single object. Now
scans for the
resultentry containingmodelUsage,total_cost_usd,num_turns, andduration_ms.Sequential /review-pr command
Replace the official
code-reviewplugin (5 parallel agents,$2.77/review) with a forked sequential version at
.claude/commands/review-pr.md.Same quality checks:
Key differences:
Expected cost reduction: parallel agents each load full
context independently (3.2M + 1.4M cache tokens last run).
Sequential reuses the cached context across checks.
🤖 Generated with Claude Code