feat(bmm): add automatic adversarial code review to quick-dev workflow
Adds Step 5 to quick-dev that automatically runs adversarial code review after implementation completes. Captures baseline commit at workflow start and reviews all changes (tracked + newly created files) using a cynical reviewer persona via subagent, CLI fallback, or inline self-review. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Alex Verkhovsky
parent
9fc0f78f7c
commit
642d5f8365
|
|
@ -23,3 +23,11 @@
|
|||
- [ ] Acceptance criteria satisfied
|
||||
- [ ] Tech-spec updated (if applicable)
|
||||
- [ ] Summary provided to user
|
||||
|
||||
## Adversarial Review
|
||||
|
||||
- [ ] Diff constructed (tracked changes from {baseline_commit} + new untracked files)
|
||||
- [ ] Adversarial review executed (subagent preferred)
|
||||
- [ ] Findings presented with severity and classification
|
||||
- [ ] User chose handling approach (walk through / auto-fix / skip)
|
||||
- [ ] Findings resolved or acknowledged
|
||||
|
|
|
|||
|
|
@ -15,6 +15,8 @@
|
|||
|
||||
<step n="1" goal="Load project context and determine execution mode">
|
||||
|
||||
<action>Record current HEAD as baseline for later review. Run `git rev-parse HEAD` and store the result as {baseline_commit}.</action>
|
||||
|
||||
<action>Check if {project_context} exists. If yes, load it - this is your foundational reference for ALL implementation decisions (patterns, conventions, architecture).</action>
|
||||
|
||||
<action>Parse user input:
|
||||
|
|
@ -170,7 +172,7 @@ Use holistic judgment, not mechanical keyword matching.</action>
|
|||
|
||||
</step>
|
||||
|
||||
<step n="4" goal="Verify and complete">
|
||||
<step n="4" goal="Verify and transition to review">
|
||||
|
||||
<action>Verify: all tasks [x], tests passing, AC satisfied, patterns followed</action>
|
||||
|
||||
|
|
@ -185,17 +187,89 @@ Use holistic judgment, not mechanical keyword matching.</action>
|
|||
**Tests:** {{test_summary}}
|
||||
**AC Status:** {{ac_status}}
|
||||
|
||||
---
|
||||
Running adversarial code review...
|
||||
</output>
|
||||
|
||||
**Before committing (Recommended): Copy this code review prompt to a different LLM**
|
||||
<action>Proceed immediately to step 5</action>
|
||||
|
||||
</step>
|
||||
|
||||
<step n="5" goal="Adversarial code review (automatic)">
|
||||
|
||||
<action>Construct diff of all changes since workflow started and capture as {diff_output}:
|
||||
|
||||
**Tracked file changes:**
|
||||
|
||||
```bash
|
||||
git diff {baseline_commit}
|
||||
```
|
||||
|
||||
**New files created by this workflow:**
|
||||
Only include untracked files that YOU actually created during steps 2-4. Do not include pre-existing untracked files. For each new file you created, include its full content as a "new file" addition.
|
||||
|
||||
Combine both into {diff_output} for review. Do NOT `git add` anything - this is read-only inspection.</action>
|
||||
|
||||
<action>Execute adversarial review using this hierarchy (try in order until one succeeds):
|
||||
|
||||
1. **Spawn subagent** (preferred) - pass the diff output along with this prompt:
|
||||
|
||||
```
|
||||
You are a cynical, jaded code reviewer with zero patience for sloppy work. These uncommitted changes were submitted by a clueless weasel and you expect to find problems. Find at least five issues to fix or improve in it. Number them. Be skeptical of everything.
|
||||
You are a cynical, jaded code reviewer with zero patience for sloppy work. This diff was submitted by a clueless weasel and you expect to find problems. Find at least five issues to fix or improve. Number them. Be skeptical of everything.
|
||||
|
||||
<diff>
|
||||
{diff_output}
|
||||
</diff>
|
||||
```
|
||||
|
||||
2. **CLI fallback** - pipe diff to `claude --print` with same prompt
|
||||
|
||||
3. **Inline self-review** - Review the diff output yourself using the cynical reviewer persona above
|
||||
</action>
|
||||
|
||||
<check if="zero findings returned">
|
||||
<action>HALT - Zero findings is suspicious. Adversarial review should always find something. Request user guidance.</action>
|
||||
</check>
|
||||
|
||||
<action>Process findings:
|
||||
|
||||
- Assign IDs: F1, F2, F3...
|
||||
- Assign severity: 🔴 Critical | 🟠 High | 🟡 Medium | 🟢 Low
|
||||
- Classify each: **real** (confirmed issue) | **noise** (false positive) | **uncertain** (needs discussion)
|
||||
</action>
|
||||
|
||||
<output>**Adversarial Review Findings**
|
||||
|
||||
| ID | Severity | Classification | Finding |
|
||||
| --- | -------- | -------------- | ------- |
|
||||
| F1 | 🟠 | real | ... |
|
||||
| F2 | 🟡 | noise | ... |
|
||||
| ... |
|
||||
|
||||
</output>
|
||||
|
||||
<action>You must explain what was implemented based on {user_skill_level}</action>
|
||||
<ask>How would you like to handle these findings?
|
||||
|
||||
**[1] Walk through** - Discuss each finding individually
|
||||
**[2] Auto-fix** - Automatically fix issues classified as "real"
|
||||
**[3] Skip** - Acknowledge and proceed to commit</ask>
|
||||
|
||||
<check if="1">
|
||||
<action>Present each finding one by one. For each, ask: fix now / skip / discuss</action>
|
||||
<action>Apply fixes as approved</action>
|
||||
</check>
|
||||
|
||||
<check if="2">
|
||||
<action>Automatically fix all findings classified as "real"</action>
|
||||
<action>Report what was fixed</action>
|
||||
</check>
|
||||
|
||||
<check if="3">
|
||||
<action>Acknowledge findings were reviewed and user chose to skip</action>
|
||||
</check>
|
||||
|
||||
<output>**Review complete. Ready to commit.**</output>
|
||||
|
||||
<action>Explain what was implemented based on {user_skill_level}</action>
|
||||
|
||||
</step>
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue