SymbX/systemd
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ci: Add back subagents and stop using --json-schema in claude-review

Browse Source 
Let's stop using --json-schema and instead have claude write a JSON
file in the repo root which we pass around as an artifact similar to
how we pass around the input. This works around the bug where claude
receives task notifications after producing structured output which
breaks the structured output.
This commit is contained in:
Daan De Meyer
2026-03-18 16:50:17 +01:00
committed by Daan De Meyer
parent 26230ed5c1
commit d105f4c5a5
1 changed files with 77 additions and 81 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/claude-review.yml
+77 -81
View File
@@ -143,9 +143,6 @@ jobs:
contents: read
id-token: write # Authenticate with AWS via OIDC
outputs:
structured_output: ${{ steps.claude.outputs.structured_output }}
steps:
- uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd
with:
@@ -180,54 +177,6 @@ jobs:
- name: Run Claude Code
id: claude
uses: anthropics/claude-code-action@26ec041249acb0a944c0a47b6c0c13f05dbc5b44
env:
REVIEW_SCHEMA: >-
{
"type": "object",
"required": ["comments", "summary"],
"properties": {
"summary": {
"type": "string",
"description": "A markdown summary of the review to post as a top-level tracking comment"
},
"comments": {
"type": "array",
"items": {
"type": "object",
"required": ["file", "severity", "body", "commit"],
"properties": {
"file": {
"type": "string",
"description": "Path to the file relative to the repo root"
},
"line": {
"type": "integer",
"description": "Line number in the diff (new file side) to attach the comment to"
},
"severity": {
"type": "string",
"enum": ["must-fix", "suggestion", "nit"]
},
"body": {
"type": "string",
"description": "The review comment body in markdown"
},
"commit": {
"type": "string",
"description": "The SHA of the PR commit that introduced the code being commented on"
}
}
}
},
"resolve": {
"type": "array",
"items": {
"type": "integer"
},
"description": "REST API IDs of existing review comments whose threads should be resolved because the issue was addressed or the author left a disagreeing reply"
}
}
}
with:
use_bedrock: "true"
# Required by claude-code-action even though Claude itself doesn't
@@ -256,26 +205,22 @@ jobs:
}
}
}
# Agent and TaskCreate are disabled because background task
# notifications race with --json-schema structured output,
# causing a redundant turn that overwrites the result.
# See https://github.com/anthropics/claude-code/issues/33872
claude_args: |
--model us.anthropic.claude-opus-4-6-v1
--max-turns 200
--disallowedTools "WebFetch,WebSearch,Agent,TaskCreate"
--disallowedTools "WebFetch,WebSearch"
--setting-sources user
--json-schema '${{ env.REVIEW_SCHEMA }}'
prompt: |
REPO: ${{ github.repository }}
PR NUMBER: ${{ needs.setup.outputs.pr_number }}
You are a code reviewer for the ${{ github.repository }} project. Review this pull request and
produce a structured JSON result containing your review. Do NOT attempt
to post comments yourself — just return the JSON. You are in the upstream repo
with the PR branch available as `pr-review`. Do not apply or merge the patch.
You have no network access — all required context has been pre-fetched
locally. You cannot spawn subagents or background tasks.
produce a JSON result containing your review and write it to
`review-result.json` in the repo root. Do NOT attempt to post
comments yourself. You are in the upstream repo with the PR branch
available as `pr-review`. Do not apply or merge the patch.
You have no network access — all required context has been
pre-fetched locally.
## Phase 1: Read context
@@ -293,15 +238,31 @@ jobs:
`git log --reverse --format=%H HEAD..pr-review` to list the PR commits, and
`git show <sha>` or `git diff <sha>~1..<sha>` to access commit diffs.
## Phase 2: Per-commit review
## Phase 2: Review commits
Review each commit in the PR sequentially. For each commit, use
`git show <sha>` or `git diff <sha>~1..<sha>` to fetch the diff,
and read relevant source files in the codebase to verify findings.
Review every commit in the PR. Use subagents to parallelize the
work — decide how many subagents to spawn and how to divide commits
between them and the main conversation based on the number and size
of commits. Very large commits can be assigned to multiple subagents
for extra thoroughness. Always review some commits yourself so you
have useful work to do while subagents run in the background. For
single-commit PRs with small diffs, just review directly without
subagents. For large single-commit PRs, have multiple subagents
review it independently to maximize coverage.
For each commit, review code quality, style, potential bugs, and
security implications. Collect issues in the format:
`{"file": "path", "line": <number> (optional), "severity": "must-fix|suggestion|nit", "body": "...", "commit": "<sha>"}`
IMPORTANT: Always spawn subagents with `isolation: "worktree"` so
each gets its own git worktree. This prevents concurrent git
operations from interfering with each other. Because worktrees
do not include untracked files, first `git add pr-context.json`
so it is available in worktrees.
Each reviewer (you or a subagent) uses `git show <sha>` or
`git diff <sha>~1..<sha>` to fetch diffs, reads `pr-context.json`
for PR context, and reads the codebase to verify findings.
Each reviewer reviews code quality, style, potential bugs, and security
implications. It must return a JSON array of issues:
`[{"file": "path", "line": <number> (optional), "severity": "must-fix|suggestion|nit", "body": "...", "commit": "<sha>"}]`
The `commit` field MUST be the SHA of the commit being reviewed. Only
comment on changes in that commit — not preceding commits.
@@ -310,7 +271,7 @@ jobs:
inside a diff hunk**. GitHub rejects lines outside the diff context. If you
cannot determine a valid diff line, omit `line`.
You MUST verify findings before including them:
Each reviewer MUST verify findings before returning them:
- For style/convention claims, check at least 3 existing examples in the
codebase to confirm the pattern actually exists before flagging a violation.
- For "use X instead of Y" suggestions, confirm X actually exists and works.
@@ -318,7 +279,7 @@ jobs:
## Phase 3: Collect, deduplicate, and summarize
After reviewing all commits:
After all reviews (yours and any subagents') are done:
1. Collect all issues. Merge duplicates (same file, lines within 3 of each other, same problem).
2. Drop low-confidence findings.
3. Check the existing inline review comments fetched in Phase 1. Do NOT include a
@@ -379,19 +340,42 @@ jobs:
not available, git commands that failed, etc.), append a `### Errors`
section to the summary listing each failed action and the error message.
## CRITICAL: Return structured JSON output
## CRITICAL: Write review result to file
Your FINAL action must be to return a JSON object matching the following
JSON schema — do NOT end with a text summary or narrative. The `--json-schema`
flag is set, so your last response must be the structured JSON result, not a
text message.
Your FINAL action must be to write `review-result.json` in the repo
root. The file must contain a JSON object with the following schema:
```json
${{ env.REVIEW_SCHEMA }}
{
"summary": "...",
"comments": [
{
"file": "path/to/file",
"line": 42,
"severity": "must-fix|suggestion|nit",
"body": "review comment in markdown",
"commit": "abc123"
}
],
"resolve": [12345]
}
```
- `summary` (required): markdown summary for the tracking comment
- `comments` (required): array of review comments; `line` is optional
- `resolve` (optional): REST API IDs of review comment threads to resolve
Do NOT attempt to post comments or use any MCP tools to modify the PR.
- name: Upload review result
if: always()
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
with:
path: review-result.json
if-no-files-found: ignore
archive: false
retention-days: 7
post:
runs-on: ubuntu-latest
needs: [setup, review]
@@ -401,15 +385,22 @@ jobs:
pull-requests: write
steps:
- name: Download review result
if: needs.review.result == 'success'
continue-on-error: true
uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c
with:
name: review-result.json
- name: Post review comments
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd
env:
STRUCTURED_OUTPUT: ${{ needs.review.outputs.structured_output }}
REVIEW_RESULT: ${{ needs.review.result }}
PR_NUMBER: ${{ needs.setup.outputs.pr_number }}
COMMENT_ID: ${{ needs.setup.outputs.comment_id }}
with:
script: |
const fs = require("fs");
const owner = context.repo.owner;
const repo = context.repo.repo;
const prNumber = parseInt(process.env.PR_NUMBER, 10);
@@ -431,9 +422,14 @@ jobs:
return;
}
/* Parse Claude's structured output. */
const raw = process.env.STRUCTURED_OUTPUT;
console.log("Structured output from Claude:");
/* Parse Claude's review result from the downloaded artifact. */
let raw = "";
try {
raw = fs.readFileSync("review-result.json", "utf8");
} catch (e) {
console.log(`Failed to read review-result.json: ${e.message}`);
}
console.log("Review result from Claude:");
console.log(raw || "(empty)");
let comments = [];