stack/packages/mosaic/framework/guides/CODE-REVIEW.md

Code Review Guide

Hard Requirement

If an agent modifies source code, code review is REQUIRED before completion. Do not mark code-change tasks done until review is completed and blockers are resolved or explicitly tracked. If code/config/API contract/auth behavior changed and required docs are missing, this is a BLOCKER. If tests pass but acceptance criteria are not verified by situational evidence, this is a BLOCKER. If implementation diverges from docs/PRD.md or docs/PRD.json without PRD updates, this is a BLOCKER.

Merge strategy enforcement (HARD RULE):

• PR target for delivery is main.
• Direct pushes to main are prohibited.
• Merge to main MUST be squash-only.
• Use ~/.config/mosaic/tools/git/pr-merge.sh -n {PR_NUMBER} -m squash --expect-head {approved_full_sha} (or PowerShell equivalent).

An estate MAY carry a documented exception for a repository whose gates are commit hooks rather than review. Such an exception belongs in that estate's own working copy of this guide, is scoped to the named repository, and is never precedent for a second one.

Do not use pr-review.sh or issue-comment.sh to post a verdict (mosaicstack#1280). Post through a direct authenticated API call as your own seat, or hand the verdict to the requesting seat. Handing it over is a legitimate delivery path, not a fallback.

Evidence Discipline (applies to every finding)

The checklist below says what to look at. This section says when you are allowed to believe what you saw. Every rule here was earned by a wrong conclusion that reached a report.

1. A finding is a claim about behavior. State the failing input, the path taken, and the wrong result. "This looks fragile" is not a finding.
2. A green check is not a result until you have shown it could go red. Run the control. A 0, an empty result, or a column of identical values with no failing counterpart is a non-result.
3. Measurement and explanation are separate sentences. Report the command and its output, then, as its own sentence, what you think it means.
4. Never widen the case you measured. If you checked one path, the finding covers one path.
5. Reproduce a reported failure before recording it, and say which tree you measured. Two correct measurements of two different trees disagree without either being wrong.
6. Verify by content on the ref that ships, never by ancestry of a local sha. A rebase mints new shas; a commit being an ancestor of something local proves nothing about the remote. Compare by digest against origin/<branch>.
7. Confidence is part of the finding. "I could not reproduce this" is a usable review comment. A confident guess is not.
8. Author is not reviewer (Gate-16). Do not review your own work, or work you shaped closely enough to be a co-author of. Say so and hand it back.

Measuring a shell suite

Each of these produced a wrong conclusion before it was written down.

9. cmd | tail; echo rc=$? reports tail's exit code, not cmd's. It reads as a pass when the command failed. Redirect to a file and check rc directly, or use ${PIPESTATUS[0]}.

10. Under set -o pipefail, a missed glob makes ls exit 2, the pipeline inherits it, and set -e kills the run. Iterate a glob with a for loop and an -e test instead of piping ls.

11. A suite that exits nonzero with ZERO output is an environment question, not a defect in the code under review. The usual cause is a sourced dependency that is absent, so set -e kills the first case before anything prints. Extract whole tool trees — tools/git alone is missing tools/_lib/credentials.sh. Isolate the variable and prove it by adding only that back.

12. git -C <dir> in a directory that is not itself a repo answers from the enclosing repo. A scratch tree under ~/.mosaic reports ~/.mosaic's HEAD, not the PR's, and every conclusion drawn from it describes the wrong tree. Confirm git rev-parse --show-toplevel is the tree you think it is before trusting any git output.

13. Run the repository's PINNED tool version. npx <tool> resolves a local node_modules install when one is present and fetches the latest release when one is not, so the same command answers differently depending on where it ran. A reviewer measuring in a fresh clone or a detached worktree — which is exactly where reviewers measure — has no node_modules and silently gets the latest release instead of the pinned one. Measured on mosaicstack#1313: the lockfile pins prettier 3.8.1, under which three guides pass; a version-less npx in a worktree resolved 3.9.6, under which the same three fail; and 3.0.0, the floor of the declared ^3.0.0 range, fails a different one. Three versions, three verdicts, identical bytes. Use node_modules/.bin/<tool>, or name the version the lockfile pins.

14. A formatter or linter declared as a range is a dated verdict, not a fact. If a lockfile pins it, the gate is reproducible today and will disagree with itself the day the pin moves. Report a formatting failure with the version that produced it, always.

Feedback Categories

• Blocker: must fix before merge (security, bugs, test failures)
• Should Fix: important but not blocking (code quality, minor issues)
• Suggestion: optional improvement (style preference, nice-to-have)
• Question: seeking clarification

Review Checklist

Reviewer seats split this checklist by class rather than duplicating it. A seat reviews its own sections in full and may raise anything it notices outside them as a Suggestion, never as a Blocker on someone else's ground.

Reviewer class	Owns
rev-code-*	1 Correctness, 3 Testing, 4 Code Quality, 4a TypeScript, 5 Documentation, 6 Performance, 7 Dependencies
rev-security-*	2 Security, 2a OWASP

Where two seats of the same class review the same change, they review independently and compare after. A second seat that reads the first seat's findings before measuring is a proofreader, not a second opinion.

1. Correctness

• Code does what the issue/PR description says
• Code aligns with active PRD requirements
• Acceptance criteria are mapped to concrete verification evidence
• Edge cases are handled
• Error conditions are managed properly
• No obvious bugs or logic errors

2. Security

• No hardcoded secrets or credentials
• Input validation at boundaries
• SQL injection prevention (parameterized queries)
• XSS prevention (output encoding)
• Authentication/authorization checks present
• Sensitive data not logged
• Secrets follow Vault structure (see docs/vault-secrets-structure.md)

2a. OWASP Coverage (Required)

• OWASP Top 10 categories were reviewed for change impact
• Access control checks verified on protected actions
• Cryptographic handling validated (keys, hashing, TLS assumptions)
• Injection risks reviewed for all untrusted inputs
• Security misconfiguration risks reviewed (headers, CORS, defaults)
• Dependency/component risk reviewed (known vulnerable components)
• Authentication/session flows reviewed for failure paths
• Logging/monitoring preserves detection without leaking sensitive data

3. Testing

• Tests exist for new functionality
• Tests cover happy path AND error cases
• Situational tests cover all impacted change surfaces (primary gate)
• Tests validate required behavior/outcomes, not only internal implementation details
• TDD was applied when required by guides/QA-TESTING.md
• Coverage meets 85% minimum
• Tests are readable and maintainable
• No flaky tests introduced

4. Code Quality

• Follows Google Style Guide for the language
• Functions are focused and reasonably sized
• No unnecessary complexity
• DRY - no significant duplication
• Clear naming for variables and functions
• No dead code or commented-out code

4a. TypeScript Strict Typing (see TYPESCRIPT.md)

• NO any types — explicit types required everywhere
• NO lazy unknown — only for error catches with immediate narrowing
• Explicit return types on all exported/public functions
• Explicit parameter types — never implicit any
• No type assertions (as Type) — use type guards instead
• No non-null assertions (!) — use proper null handling
• Interfaces for objects — not inline types
• Discriminated unions for variant types
• DTO files used at boundaries — module/API contracts are in *.dto.ts, not inline payload types

5. Documentation

• Complex logic has explanatory comments
• Required docs updated per guides/DOCUMENTATION.md
• Public APIs are documented
• Private/internal APIs are documented
• API input/output schemas are documented
• API permissions/auth requirements are documented
• Site map updates are present when navigation changed
• README updated if needed
• Breaking changes noted

6. Performance

• No obvious N+1 queries
• No blocking operations in hot paths
• Resource cleanup (connections, file handles)
• Reasonable memory usage

7. Dependencies

• No deprecated packages
• No unnecessary new dependencies
• Dependency versions pinned appropriately

Review Process

Use ~/.config/mosaic/templates/docs/DOCUMENTATION-CHECKLIST.md whenever code/API/auth/infra changes are present.

Getting Context

# List the issue being addressed
~/.config/mosaic/tools/git/issue-list.sh -i {issue-number}

# View the changes
git diff main...HEAD

Providing Feedback

• Be specific: point to exact lines/files
• Explain WHY something is problematic
• Suggest alternatives when possible
• Distinguish between blocking issues and suggestions
• Be constructive, not critical of the person

Review Comment Format

[BLOCKER] Line 42: SQL injection vulnerability
The user input is directly interpolated into the query.
Use parameterized queries instead:
`db.query("SELECT * FROM users WHERE id = ?", [userId])`

[SUGGESTION] Line 78: Consider extracting to helper
This pattern appears in 3 places. A shared helper would reduce duplication.

After Review

1. Update issue with review status
2. If changes requested, assign back to author
3. If approved, note approval in issue comments
4. For merges, ensure CI passes first
5. Merge PR to main with squash strategy only