Skip to content
Agentmart
Submit a tool

ultra-review

Parallel code review orchestration with 7 specialized agents + coordinator.

Skill by rocky2431

IndexedMIT★ 11
Write a review

Chalk Labs

Need help deploying ultra-review?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/rocky2431-ultra-review

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Parallel code review orchestration with 7 specialized agents + coordinator. Zero context pollution - all output via JSON files.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found
Runs as
Python package

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/ultra-review

What's inside

Skill name
ultra-review
Description
Parallel code review orchestration with 7 specialized agents + coordinator. Zero context pollution - all output via JSON files.
License
MIT
Instructions
435 lines in SKILL.md · view on GitHub

Before installing

  • Ships executable scripts

Files (4)

  • SKILL.md
  • unified-schema.md · references/
  • review_verdict_update.py · scripts/
  • review_wait.py · scripts/

SKILL.md (first lines)

# /ultra-review - Ultra Review System

Orchestrates parallel code review using specialized agents. All findings written to JSON files to prevent context window pollution.

## Workflow Tracking (MANDATORY)

**On command start**, create tasks for each major step using `TaskCreate`:

| Step | Subject | activeForm |
|------|---------|------------|
| 1 | Setup & Scope Detection | Detecting diff scope... |
| 2 | Agent Selection | Selecting review agents... |
| 3 | Background Execution | Running review agents... |
| 4 | Wait & Coordinate | Waiting for agent results... |
| 5 | Report to User | Generating review report... |

**Before each step**: `TaskUpdate` → `status: "in_progress"`
**After each step**: `TaskUpdate` → `status: "completed"`
**On context recovery**: `TaskList` → resume from last incomplete step

## Usage

```
/ultra-review              # Full review (smart skip based on diff content)
/ultra-review all          # Force ALL 6 agents, no auto-skip (pre-merge gate)
/ultra-review quick        # Quick review (review-code only)
/ultra-review security     # Security focus (review-code + review-errors)
/ultra-review tests        # Test quality focus (review-tests only)
/ultra-review recheck      # Re-check only P0/P1 files from last session
/ultra-review delta        # Review only changes since last review session
/ultra-review <agents...>  # Custom: e.g., /ultra-review tests errors types
```

### Scope Options (combinable with any mode)

```
/ultra-review --pr 123            # Review PR #123 diff (base...head)
/ultra-review --range main..HEAD  # Review specific commit range
/ultra-review --range abc123      # Review single commit
/ultra-review security --pr 42    # Security review scoped to PR #42
```

## Orchestration Flow

### Phase 1: Setup & Scope Detection

```python
# 1. Determine session context
BRANCH = git branch --show-current            # e.g., "feat/task-3-auth"
ITER = count existing sessions matching this branch + 1  # iteration number

# 2. Create session directory with full context
SESSION_ID = "<YYYYMMDD-HHmmss>-<branch>-iter<N>"
# e.g., "20260214-103000-feat-task-3-auth-iter2"
SESSION_PATH = ".ultra/reviews/{SESSION_ID}/"
mkdir -p SESSION_PATH

# 3. Update session index (placeholder — verdict filled in Phase 5)
# Append entry to .ultra/reviews/index.json with verdict="pending"
# See Session Index below for full schema

# 4. Cleanup (see Lifecycle Management below)

# 5. Detect diff scope (priority order)
```

**Scope resolution priority:**

| Flag | Diff Command | DIFF_RANGE |
|------|-------------|------------|
| `--pr NUMBER` | `gh pr diff NUMBER --name-only` | PR base...head |
| `--range RANGE` | `git diff RANGE --name-only` | Specified range |
| _(default)_ | `git diff --name-only HEAD` + `git diff --cached --name-only` | HEAD (unstaged + staged) |

**PR scope details** (`--pr NUMBER`):
```bash
# Get PR metadata for session naming
gh pr view NUMBER --json headRefName,baseRefName,title
# Get changed files
gh pr diff NUMBER --name-only
# Set DIFF_RANGE to base...head for agents
DIFF_RANGE="$(gh pr view NUMBER --json baseRefName -q .baseRefName)..HEAD"
```

**Range scope details** (`--range RANGE`):
```bash
git diff RANGE --name-only
DIFF_RANGE="RANGE"
```

### Phase 2: Agent Selection

Based on the mode argument and file classification:

**Mode: `full` (default)**
Select all applicable agents based on diff content:

| Agent | Skip Condition | Warning |
|-------|---------------|---------|
| review-code | Never skip | - |
| review-tests | No test files in diff AND no source files without tests | WARN: "Code changed without test updates" |
| review-errors | No error handling code detected | - |
| review-design | No type definitions AND no function-level changes in diff | - |
| review-comments | No comment changes in diff | - |
| review-ac-drift | No `.ultra/tasks/contexts/` directory OR no active task locatable | - |

**Auto-skip detection commands:**
```bash
# Has type definitions?
grep -l "interface \|type \|class \|enum " <diff_files>

# Has comment changes?
git diff HEAD -- <diff_files> | grep "^[+-].*//\|^[+-].*\*\|^[+-].*#"

# Has test files?
echo "<diff_files>" | grep -E "\.(test|spec)\.(ts|tsx|js|jsx)$|test_.*\.py$|.*_test\.go$"

# Has error handling?
grep -l "try\|catch\|\.catch\|throw\|Error(" <diff_files>

From rocky2431/ultra-builder-pro, MIT.

More from rocky2431/ultra-builder-pro

Signals

Stars
11
Forks
4
Open issues
0
Last commit (this folder)
Apr 30, 2026 (6 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from rocky2431

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get ultra-review running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying ultra-review?

CallGet help
Cite this page

ultra-review skill. Agent Mart. https://agentmart.dev/skills/rocky2431-ultra-review (last updated 2026-10-24)