Technical Report Pro

Generate professional, structured technical documentation following engineering best practices.

Report Types

Type	Template	Use Case
Architecture Review	`arch-review.md`	System design analysis, component relationships
Post-Mortem	`post-mortem.md`	Incident analysis, root cause, remediation
Research Summary	`research-summary.md`	Findings synthesis, recommendations
Debug Investigation	`debug-investigation.md`	Problem diagnosis, solution verification

Standard Structure (ARCHIE-ENG-001)

Every report MUST include:

1. Executive Summary (≤200 words)

•What: One-sentence description of the subject
•Why: Business/technical impact
•Outcome: Key findings or decisions
•Action: Next steps (if applicable)

2. Technical Justification ("The Why")

•Problem statement with context
•Constraints and requirements
•Alternative approaches considered
•Rationale for chosen approach

3. Implementation Details ("The How")

•Architecture diagrams (ASCII or Mermaid)
•Component breakdown
•Data flows
•Configuration details
•Code references (file paths, line numbers)

4. Verification Proofs ("The Evidence")

•Test results with actual output
•Metrics (before/after)
•Logs or command output
•Screenshots or recordings (paths)

Usage

Quick Generation

bash

# Generate from template
python scripts/report_gen.py --type arch-review --title "OpenClaw Hive Architecture"

# With context gathering
python scripts/report_gen.py --type post-mortem --scan research/ --title "BlueBubbles Integration Issue"

Template Selection

Choose template based on:

Trigger	Template
"How does X work?"	arch-review
"What went wrong?"	post-mortem
"What did we learn?"	research-summary
"Why is X broken?"	debug-investigation

Writing Guidelines

•Language: English only (per protocol)
•Tone: Technical, precise, no fluff
•Evidence: Every claim needs proof
•Brevity: Prefer tables/lists over paragraphs
•Actionable: End with clear next steps

Context Gathering

Before writing, scan for context:

bash

# Find relevant files
grep -r "keyword" research/ dev/ --include="*.md"

# Check recent work
ls -lt research/ | head -10

# Load memory
cat memory/observations.md | grep -i "topic"

Templates

Templates are in assets/templates/. Load with:

python

from pathlib import Path
template = Path("assets/templates/arch-review.md").read_text()

Example Output

See references/examples/ for sample reports demonstrating each template.