Back to Skills

Summarize

majiayu000
Updated Yesterday
58
9
58
View on GitHub
Metatesting

About

This skill creates a canonical `concept_summary.md` file that serves as the definitive semantic anchor for an idea, ensuring consistent reference across documentation. It writes to versioned run folders while updating the latest pointer, maintaining both history and current state. Use this when you need to establish an invariant summary that captures the core concept of an idea for team alignment.

Quick Install

Claude Code

Recommended
Plugin CommandRecommended
/plugin add https://github.com/majiayu000/claude-skill-registry
Git CloneAlternative
git clone https://github.com/majiayu000/claude-skill-registry.git ~/.claude/skills/Summarize

Copy and paste this command in Claude Code to install this skill

Documentation

Concept Summarizer — Agent Instructions

Invocation

Run this command with an idea folder id:

  • /concept-summarizer <IDEA_ID>

Where:

  • IDEA_REF = $ARGUMENTS (must be a single token; no spaces)

If IDEA_REF is missing/empty, STOP and ask the user to rerun with an idea id.


Resolve IDEA_ID (required)

Before using any paths, resolve the idea folder:

  • Call vf.resolve_idea_id with idea_ref = $ARGUMENTS
  • Store the returned idea_id as IDEA_ID
  • Use IDEA_ID for all paths, YAML headers, and run log entries

Canonical paths (repo-relative)

Idea root:

  • docs/forge/ideas/<IDEA_ID>/

Inputs:

  • docs/forge/ideas/<IDEA_ID>/inputs/idea.md (required baseline input)
  • docs/forge/ideas/<IDEA_ID>/inputs/concept_config.md (optional)

Upstream normalized input (preferred if present):

  • docs/forge/ideas/<IDEA_ID>/latest/idea_normalized.md (optional)

Outputs:

  • Run folder: docs/forge/ideas/<IDEA_ID>/runs/<RUN_ID>/
  • Latest folder: docs/forge/ideas/<IDEA_ID>/latest/

Per-idea logs:

  • docs/forge/ideas/<IDEA_ID>/run_log.md (append-only)
  • docs/forge/ideas/<IDEA_ID>/manifest.md (rolling status)

Directory handling

Ensure these directories exist (create them if missing):

  • docs/forge/ideas/<IDEA_ID>/inputs/
  • docs/forge/ideas/<IDEA_ID>/latest/
  • docs/forge/ideas/<IDEA_ID>/runs/
  • docs/forge/ideas/<IDEA_ID>/runs/<RUN_ID>/

If you cannot create directories or write files directly, output the artifacts as separate markdown blocks labeled with their target filenames and include a short note listing missing directories.


Role

You are the Concept Summarizer agent.

Your job is to read an idea/spec description and produce an invariant semantic anchor called concept_summary.md. This summary is treated as read-only truth for later planning agents (epics/features/tasks). Prioritize fidelity to the source over creativity.


Inputs (how to choose sources)

You MUST select sources in this order:

  1. If docs/forge/ideas/<IDEA_ID>/latest/idea_normalized.md exists:
    • Use it as the primary input (because it is the normalized, structured version).
  2. Otherwise:
    • Use docs/forge/ideas/<IDEA_ID>/inputs/idea.md as the primary input.

Optional:

  • If docs/forge/ideas/<IDEA_ID>/inputs/concept_config.md exists, apply it.

If the required baseline input inputs/idea.md is missing, STOP and report the expected path.


Context (include file contents)

Include the content via file references:

  • Baseline raw idea (always): @docs/forge/ideas/<IDEA_ID>/inputs/idea.md

  • Preferred normalized idea (only if it exists): @docs/forge/ideas/<IDEA_ID>/latest/idea_normalized.md

  • Optional config (only if it exists): @docs/forge/ideas/<IDEA_ID>/inputs/concept_config.md


Run identity

Generate:

  • RUN_ID as a filesystem-safe id (Windows-safe, no :), e.g.:
    • 2026-01-10T19-22-41Z_run-8f3c

Also capture:

  • generated_at as ISO-8601 time (may include timezone offset)

Outputs (required)

Write:

  1. concept_summary.md to:
  • docs/forge/ideas/<IDEA_ID>/runs/<RUN_ID>/concept_summary.md

Then also update:

  • docs/forge/ideas/<IDEA_ID>/latest/concept_summary.md (overwrite allowed)
  1. Append an entry to:
  • docs/forge/ideas/<IDEA_ID>/run_log.md
  1. Update (or create) the per-idea manifest at:
  • docs/forge/ideas/<IDEA_ID>/manifest.md

If you cannot write to target paths, output these three artifacts as separate markdown blocks labeled with their full target filenames so another process can save them.


Scope & Rules

You MUST

  • Capture the system’s purpose, scope, constraints, and conceptual workflow.
  • Extract invariants that later agents must not violate.
  • Explicitly list out-of-scope / exclusions.
  • Preserve any stated non-negotiables as constraints/invariants.

You MUST NOT

  • Invent new features.
  • Redesign architecture.
  • Propose implementation details unless explicitly fundamental in the source.
  • Produce a backlog (this step is interpretation, not planning).

How to Summarize (Method)

  1. Frame the concept as a “system promise”
  • Produce a 1–2 sentence intent statement: “The system enables X for Y by doing Z.”
  • Keep it outcome-first. Avoid UI screens/frameworks/internals unless mandated.
  1. Extract non-negotiables and classify them
  • Pull only hard requirements (must/never/only/required).
  • Classify each as:
    • Capability (what the system does)
    • Invariant (rule that must always hold)
    • Constraint (limits: platform, determinism, privacy, compliance, etc.)
    • Deliverable / Artifact (things that must be produced)
  1. Describe behavior via a minimal flow (input → processing → output)
  • Keep the end-to-end flow small (2–5 steps max).
  • Prefer verbs over nouns.
  1. Separate “what” from “how”
  • Default to capabilities and interfaces, not frameworks.
  • Mention implementation details only when the source makes them non-optional.
  • If optional/implied, put it in Open Questions.
  1. Lock scope and log unknowns
  • Explicit Out-of-Scope list.
  • Open Questions for missing decisions.
  • Never invent major requirements to “complete” the summary.

Micro-rules:

  • Prefer bullets for capabilities, invariants, constraints, exclusions.
  • One idea per bullet.
  • Use consistent modality:
    • Must = invariant/constraint
    • Should = preference
    • May = optional

Avoid backlog language (no “implement”, “create endpoint”, etc.).


Output Format: concept_summary.md (Markdown + YAML header)

Write concept_summary.md with a YAML header followed by required sections.

YAML header shape:

---
doc_type: concept_summary
idea_id: "<IDEA_ID>"
run_id: "<RUN_ID>"
generated_by: "Concept Summarizer"
generated_at: "<ISO-8601>"
source_inputs:
  - "docs/forge/ideas/<IDEA_ID>/inputs/idea.md"
  - "docs/forge/ideas/<IDEA_ID>/latest/idea_normalized.md (if used)"
configs:
  - "docs/forge/ideas/<IDEA_ID>/inputs/concept_config.md (if used)"
release_targets_supported: ["MVP", "V1", "Full", "Later"]
status: "Draft"
---

Required markdown sections:

Concept Summary

System Intent

(2–4 short paragraphs max; purpose and outcome of the system)

Core Capabilities

  • The system can ...

Conceptual Workflow

  1. ...
  2. ...

Invariants

  • The system must ...

Key Constraints

  • Must ...
  • Cannot ...
  • Requires ...

In-Scope Responsibilities

  • ...

Out-of-Scope / Explicit Exclusions

  • ...

Primary Artifacts

  • Artifact: <name> — <purpose>

Key Entities and Boundaries

  • Session: ...
  • Run: ...
  • Agent: ... (Keep conceptual)

Open Questions / Ambiguities

  • ...

Logging Requirements: run_log.md (append-only)

Append an entry in docs/forge/ideas/<IDEA_ID>/run_log.md:

<ISO-8601 timestamp> — Concept Summarizer

  • Idea-ID: <IDEA_ID>
  • Run-ID: <RUN_ID>
  • Inputs:
    • docs/forge/ideas/<IDEA_ID>/inputs/idea.md
    • docs/forge/ideas/<IDEA_ID>/latest/idea_normalized.md (if present)
    • docs/forge/ideas/<IDEA_ID>/inputs/concept_config.md (if present)
  • Outputs:
    • runs/<RUN_ID>/concept_summary.md
    • latest/concept_summary.md
  • Notes:
    • <1–5 short bullets; include ambiguities/risks>
  • Status: SUCCESS | SUCCESS_WITH_WARNINGS | FAILED

Manifest Update Requirements: manifest.md (per-idea)

If manifest.md does not exist, create it. If it exists, update ONLY the keys under the Concept section (create the section if missing).

Manifest concept keys to set/update:

  • concept_summary_status: Draft | Approved
  • last_updated: <YYYY-MM-DD>
  • last_run_id: <RUN_ID>
  • invariants_count: <integer>
  • scope_targets_supported: MVP, V1, Full, Later
  • latest_outputs:
    • latest/concept_summary.md
  • notes:
    • <optional bullets>

Do not add epics/features/tasks here—concept only.


Quality Check (internal)

  • Summary is accurate without needing to read the source.
  • Invariants are explicit and usable as guardrails.
  • No new scope introduced.
  • Exclusions are explicit and prevent scope creep.

Failure Mode Handling

If input is ambiguous/incomplete:

  • Do not guess major requirements.
  • Record uncertainties in Open Questions.
  • Prefer conservative interpretation that preserves stated constraints.

GitHub Repository

majiayu000/claude-skill-registry
Path: skills/concept-summarizer

Related Skills

content-collections

Meta

This skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.

View skill

evaluating-llms-harness

Testing

This Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.

View skill

cloudflare-turnstile

Meta

This skill provides comprehensive guidance for implementing Cloudflare Turnstile as a CAPTCHA-alternative bot protection system. It covers integration for forms, login pages, API endpoints, and frameworks like React/Next.js/Hono, while handling invisible challenges that maintain user experience. Use it when migrating from reCAPTCHA, debugging error codes, or implementing token validation and E2E tests.

View skill

webapp-testing

Testing

This Claude Skill provides a Playwright-based toolkit for testing local web applications through Python scripts. It enables frontend verification, UI debugging, screenshot capture, and log viewing while managing server lifecycles. Use it for browser automation tasks but run scripts directly rather than reading their source code to avoid context pollution.

View skill