$npx -y skills add evanklem/evanflow --skill evanflow-reviewCode review — both giving and receiving feedback. Verifying observed behavior over performative agreement. Use when reviewing a PR, requesting review on completed work, or processing review feedback you received.
| 1 | # EvanFlow: Review |
| 2 | |
| 3 | |
| 4 | ## Vocabulary |
| 5 | |
| 6 | See `evanflow` meta-skill. |
| 7 | |
| 8 | ## When to Use |
| 9 | |
| 10 | - Completed work, want a review pass before integration |
| 11 | - Reviewing someone else's PR |
| 12 | - Received review feedback and need to process it (especially if you disagree) |
| 13 | |
| 14 | ## Two Halves |
| 15 | |
| 16 | ### A. Requesting Review (giving work to be reviewed) |
| 17 | |
| 18 | #### 1. Self-review first |
| 19 | |
| 20 | Before asking anyone to look: |
| 21 | |
| 22 | - Run the project's quality checks (`tsc`, `lint`, `test`) |
| 23 | - Re-read the diff with fresh eyes |
| 24 | - Check the diff matches the spec/plan — no scope creep |
| 25 | - Look for: leftover console.logs, commented-out code, unused imports, secrets accidentally committed |
| 26 | - Apply the deletion test to new modules: do they earn their existence? |
| 27 | |
| 28 | #### 2. Frame the review request |
| 29 | |
| 30 | State explicitly: |
| 31 | |
| 32 | - **Goal**: what behavior changed |
| 33 | - **Approach**: why this approach (link to spec/plan) |
| 34 | - **Risk areas**: where you most want a critical eye |
| 35 | - **Verification done**: what you tested locally; what's untested |
| 36 | - **Open questions**: things you're unsure about |
| 37 | |
| 38 | A reviewer can't help you find what you didn't flag. |
| 39 | |
| 40 | ### B. Receiving Review |
| 41 | |
| 42 | #### 1. Read all feedback first |
| 43 | |
| 44 | Don't start fixing the first comment until you've read every comment. Patterns matter — three comments on similar code might mean a structural issue, not three separate fixes. |
| 45 | |
| 46 | #### 2. Embedded Grill — Challenge feedback you disagree with |
| 47 | |
| 48 | For any feedback that feels wrong: |
| 49 | |
| 50 | - "Is the reviewer working from a wrong assumption? What context might they be missing?" |
| 51 | - "Does the project's CLAUDE.md or `docs/adr/` already address this?" |
| 52 | - "Would the suggested change make a different test pass that currently doesn't exist?" |
| 53 | - "Is this a stylistic preference or a correctness issue?" |
| 54 | |
| 55 | If after grilling you still disagree, **respond with reasoning, not capitulation.** Performative agreement that ships bad code is worse than respectful disagreement. |
| 56 | |
| 57 | #### 3. Group changes |
| 58 | |
| 59 | Batch similar feedback into single edits. Don't make 17 separate one-line changes. |
| 60 | |
| 61 | #### 4. Verify after each batch |
| 62 | |
| 63 | Run the quality checks. Don't push a "review fixes" commit that breaks the build. |
| 64 | |
| 65 | ## Hard Rules |
| 66 | |
| 67 | - **Verify before claiming a fix is done.** Run the test that the reviewer flagged. |
| 68 | - **Don't capitulate to feedback you can't justify.** If you can't articulate WHY the change is right, don't make it. |
| 69 | - **Don't agree-and-defer.** "Good catch, will fix later" with no follow-up is debt. |
| 70 | - **Never auto-commit, never auto-stage, never auto-finish.** When review fixes are done, report what was done and STOP. The user decides when to commit. |
| 71 | |
| 72 | ## Hand-offs |
| 73 | |
| 74 | - Review accepted, changes complete → **report what was done and STOP. Await user direction.** |
| 75 | - Review surfaced a deeper issue → `evanflow-improve-architecture` or `evanflow-debug` |
| 76 | - Review surfaced wrong assumptions about intent → `evanflow-brainstorming` to realign |