Executive Summary Writer

executive-summary-writer

Purpose: Take a full report or document and synthesize it into a 150-300 word executive summary that a decision-maker can act on without reading the full document.

Input Schema

| Field | Type | Required | |---|---|---| | source_title | string | yes | | source_content | string | yes - full document or long excerpt | | audience | string | yes | | key_findings | string[] | yes - top 3-5 findings to surface | | recommended_actions | string[] | yes | | preserve_facts | string[] | yes |

{
  "source_title": "Q1 2024 Platform Engineering Report",
  "source_content": "...",
  "audience": "VP Engineering",
  "key_findings": ["CI build time improved 60%", "On-call volume down 40%", "2 critical infra migrations completed"],
  "recommended_actions": ["Approve Q2 headcount for observability", "Extend Kubernetes rollout to remaining 12 services"],
  "preserve_facts": ["60%", "40%", "12 services", "Q1 2024"]
}

Output Schema

{
  "title": "Executive Summary: Q1 2024 Platform Engineering",
  "summary": "...",
  "findings": ["..."],
  "recommended_actions": ["..."],
  "word_count": 215
}

Prompt Flow

Pass 1: Three-part structure: What happened (findings), What it means (interpretation in 1-2 sentences), What's next (recommended actions). Remove: P1, P3, P4, P5, P6, P7, P14, P15, P22, P23, P24. Preserve all facts.

Pass 2: "What in this summary is padding?" -> cut aggressively to under 300 words. Every sentence must do work.

Examples

Short - findings sentence

Before: "The platform team has made significant strides, showcasing their commitment to fostering a culture of continuous improvement that underscores their vital role in the organization's evolving landscape." After: "CI build time dropped 60% in Q1; on-call volume fell 40%."

Medium - full summary

Before: "In conclusion, the future looks bright for our platform infrastructure. Exciting times lie ahead as we continue this journey toward excellence and scalability." After: "Q1 work shipped two critical migrations and cut on-call load nearly in half. Q2 priority: extend Kubernetes to the remaining 12 services and hire 2 observability engineers to avoid repeating the January monitoring gap."

Long

Unit Tests

# tests/skills/test_executive_summary_writer.py
from ai_pattern_scrubber import detect_patterns

FINDING = "CI build time dropped 60% in Q1; on-call volume fell 40%."
CLOSE = "Q2 priority: extend Kubernetes to 12 services and hire 2 observability engineers."

def test_finding_no_significance_inflation():
    assert not [h for h in detect_patterns(FINDING) if h.id == 1]

def test_close_no_generic_conclusion():
    assert not [h for h in detect_patterns(CLOSE) if h.id == 24]

def test_finding_no_high_severity():
    assert not [h for h in detect_patterns(FINDING) if h.severity == "high"]

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: nicholashidalgo
• Source: nicholashidalgo/claude-skillforge
• License: MIT

Install and usage instructions live in the source repository linked above.