fix(ci): hand review diff to agent as file, not argv
PR Check / Dockerfile lint (droast) (pull_request) Skipped
PR Check / Security scan (Snyk) (pull_request) Skipped
PR Check / ESLint (pull_request) Successful in 35s
PR Check / Typecheck (pull_request) Successful in 38s
PR Check / Secret scan (gitleaks) (pull_request) Successful in 5s
PR Check / Dependency audit (pull_request) Successful in 9s
PR Check / AI code review (opencode) (pull_request) Failing after 11s

Passing the PR diff as a command-line argument exceeded
MAX_ARG_STRLEN (128KiB per single argument; the diff was 190KB,
mostly pnpm-lock.yaml). The shell failed with E2BIG ('Argument list
too long') and the tee pipeline masked the failure, so the raw error
log got posted as a 'review' comment.

- exclude pnpm-lock.yaml from the review diff (pnpm audit covers deps)
- write the diff to .pr-review-diff.patch in the workspace; the agent
  reads it with file tools
- redirect output instead of piping through tee and fail the job when
  opencode fails, so errors are never posted as review comments
This commit is contained in:
Noel Sabia committed 2026-10-18 16:25:30 +02:00
1 parent 99f6e28ad0
commit 16b9291a51
1 file changed
+23 -6
+23 -6
View File
@@ -190,9 +190,15 @@ jobs:
- name: Get PR diff - name: Get PR diff
run: | run: |
git diff --merge-base \ # lockfile changes bloat the diff (175KB here) and are pointless to
"origin/${{ github.base_ref }}" > /tmp/pr.diff # review - pnpm audit already covers dependencies. Excluding the
wc -l /tmp/pr.diff # lockfile also keeps the diff small enough to hand to the model.
# The diff goes to a file IN the workspace so the agent can read it
# with its file tools (passing 190KB as a CLI argument fails with
# "Argument list too long", MAX_ARG_STRLEN is 128KiB).
git diff --merge-base "origin/${{ github.base_ref }}" \
-- . ':(exclude)pnpm-lock.yaml' > .pr-review-diff.patch
wc -c .pr-review-diff.patch
- name: Run opencode review - name: Run opencode review
env: env:
@@ -201,12 +207,18 @@ jobs:
# deepseek-v4-pro#max: flagship reasoning model at maximum thinking # deepseek-v4-pro#max: flagship reasoning model at maximum thinking
# budget. Requires funds on the opencode account. # budget. Requires funds on the opencode account.
# Cheaper fallbacks: opencode/deepseek-v4.1-flash, or remove "#max". # Cheaper fallbacks: opencode/deepseek-v4.1-flash, or remove "#max".
# Output is redirected (not piped through tee): if opencode fails,
# the job must fail here so the error log is never posted as a
# "review" comment.
opencode run \ opencode run \
--model opencode/deepseek-v4-pro#max \ --model opencode/deepseek-v4-pro#max \
"You are a principal engineer performing a rigorous code review on a "You are a principal engineer performing a rigorous code review on a
pull request. You have tool access: read repo files for context when pull request. You have tool access: read repo files for context when
the diff alone is ambiguous, but review ONLY code changed in this PR. the diff alone is ambiguous, but review ONLY code changed in this PR.
The full PR diff is in the file .pr-review-diff.patch in the repo
root. Read it first with your file tools.
## Project context ## Project context
- Stack: Nuxt 4 (Vue 3, TypeScript) SSR web app in the repo root. - Stack: Nuxt 4 (Vue 3, TypeScript) SSR web app in the repo root.
Backend/API services are being added in Golang. Backend/API services are being added in Golang.
@@ -251,7 +263,7 @@ jobs:
unoptimized images, blocking Nitro middleware). unoptimized images, blocking Nitro middleware).
## Output format (strict) ## Output format (strict)
Review the diff at the end of this message. For every finding: Review the diff in .pr-review-diff.patch. For every finding:
[BLOCKER] file:line - what and why + concrete fix suggestion [BLOCKER] file:line - what and why + concrete fix suggestion
[WARN] file:line - what and why + concrete fix suggestion [WARN] file:line - what and why + concrete fix suggestion
[NIT] file:line - optional improvement [NIT] file:line - optional improvement
@@ -266,8 +278,13 @@ jobs:
End your response with a final verdict line, exactly one of: End your response with a final verdict line, exactly one of:
VERDICT: APPROVE (no BLOCKER findings) VERDICT: APPROVE (no BLOCKER findings)
VERDICT: REQUEST_CHANGES (at least one BLOCKER)" \ VERDICT: REQUEST_CHANGES (at least one BLOCKER)" \
"$(cat /tmp/pr.diff)" \ > /tmp/review.md 2>&1
2>&1 | tee /tmp/review.md EXIT_CODE=$?
cat /tmp/review.md
if [ $EXIT_CODE -ne 0 ]; then
echo "::error::opencode review failed (exit ${EXIT_CODE}) - NOT posting this output as a review comment"
exit $EXIT_CODE
fi
- name: Post review as PR comment - name: Post review as PR comment
continue-on-error: true # token missing/expired must not fail the build continue-on-error: true # token missing/expired must not fail the build