rclab-review

name	rclab-review
description	Generate solo synthesis/review reports from source docs — 1+ independent agents each read the same sources and write their own report. No coordination between agents.
argument-hint	<doc(s)> --agents <type1[,type2,...]> [--session <id>] [--output <path>] [--context <text>]

rclab-review

--help

If $ARGUMENTS contains --help or -h (or is empty and the user seems confused), read and display .claude/rclab-help.md, then stop. Do not proceed with any other phase.

---

Solo synthesis. 1+ agents independently read source documents and write their own report. No coordination between agents — each produces its own file from the same inputs.

For 2-agent iterative workshops on a shared document, use /rclab-workshop.

Usage

# One agent synthesizes source docs
/rclab-review session-63-W6*.md --agents <domain-generalist>

# Three agents each independently write a report
/rclab-review session-63*.md --agents skeptic,calculator,workhorse-1 --session 63

# With focus topics
/rclab-review session-74*.md --agents observer,bridge --context core claims, robustness

---

Phase 0: Parse & Validate

0a. Extract Arguments

Parse $ARGUMENTS:

Arg	Required	Default	Notes
[doc(s)]	yes (1+)	—	Source doc paths or globs (positional, before flags)
--agents	yes	—	Comma-separated agent types or short names (1 or more)
--session	no	auto-detect	Session ID (e.g., 63)
--output	no	auto-detect	Output path prefix (per-agent files generated)
--context	no	—	Focus topics or instructions passed to agents

0b. Validate

1. Source docs: Glob-resolve paths. Read 1 line of each to verify existence. Report missing and stop.
2. Agent types: Resolve short names via .claude/templates/agent-roster.md. If invalid, list available types and stop.

0c. Defaults

Session ID (if not provided):

• Extract from first source doc filename: regex session-(\d+)

Output path (if not provided, per agent):

• sessions/session-{id}/session-{id}-{short-name}-synthesis.md

If session ID unresolvable, ask the user.

---

Phase 1: Collision Check

If any output file already exists, ask: "Output file exists at {path}. Overwrite / New name / Cancel?"

---

Phase 2: Execute

For each agent in --agents, spawn a background agent in parallel:

• subagent_type: the agent type
• run_in_background: true
• name: review-{short-name}
• mode: acceptEdits

Agent prompt:

You are writing a synthesis report for the {{PROJECT_NAME}} project.

## Source Documents (read ALL of these FIRST)
{numbered list of source doc paths}

Also read your agent memory: `.claude/agent-memory/{your-type}/MEMORY.md`

{If --context provided:}
## Focus
{context text}

## Your Task

Read all source documents, then write your synthesis to: `{output_path}`

## Document Structure

Follow the template in `.claude/templates/synthesis.md`.

## Rules
- Gate verdicts from source docs are authoritative — do not re-adjudicate.
- If sources conflict, flag the conflict explicitly.
- Write ONLY the output file.
- Use the project's framing vocabulary (see the framework-framing rule if one is installed by the selected discipline pack).
- Every equation dimensionally consistent. Every approximation states its regime.

If multiple agents: create a TaskCreate per agent for progress tracking.

---

Phase 3: Verify & Report

=== RCLAB-REVIEW COMPLETE ===
Agent(s): {list}
Output: {path(s)} ({lines} lines each)
Source documents: {N}

---

Rules

1. Never overwrite files without user confirmation (collision check).
2. Never execute computations — review reports only.
3. Never re-adjudicate gate verdicts — source doc verdicts are authoritative.
4. No cross-agent coordination — each agent writes independently. For iterative 2-agent exchanges, use /rclab-workshop.
5. Never initiate shutdown — user decides.
6. Project-specific framing — if the selected discipline pack installed a framework-framing rule, it applies.

Error Handling

Condition	Action
No source docs	Show usage and stop
Source doc missing	Report which, stop
No --agents	Show usage and stop
Agent type invalid	List available types, stop
Output collision	Ask: overwrite / rename / cancel
Agent fails to produce output	Report, suggest different agent