# Edico

> Core Knowledge Management Skill. Reduces research redundancy and provides autonomous long-term memory by persisting synthesized web data.

- Skill: `xenitv1/edico` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add xenitv1/edico`
- Raw SKILL.md: https://api.skillmd.com/api/skills/xenitv1/edico/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: xenitv1 (https://skillmd.com/u/xenitv1)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/xenitv1/edico

---

<domain_overview>
# 📚 EDICO: PERSISTENCE & REDUNDANCY REDUCTION
> **MISSION:** To stop redundant web searches by the AI and transform learned insights into a permanent, collective local memory.
> **Philosophy:** Knowledge should be cumulative, not ephemeral. Edico is the bridge between conversations, ensuring that once a topic is researched, it remains accessible forever.
</domain_overview>

<iron_laws>
## 🚨 IRON LAWS
```
1. CHECK BEFORE SEARCH - Always scan the local .webdata directory BEFORE starting any deep web research.
2. NO REDUNDANCY - If existing data is fresh (less than 2 months old), use it as the primary source instead of performing a new search.
3. SOURCE-BASED TIMESTAMPS - Do NOT use the current date for the record. Use the date mentioned in the sources (e.g., article publication date) to ensure historical accuracy.
4. EXHAUSTIVE ANALYSIS - Never write a "summary." Write a COMPREHENSIVE analysis. It must be so detailed that any subsequent AI reading it will NOT need to perform a new search. Include all data points, technical nuances, and context.
5. ENGLISH ONLY PERSISTENCE - All data saved to the local database (topic, detailed_analysis, tags) MUST be written in English, regardless of the user's language or the source's language. This ensures 100% interoperability across systems.
6. AUTONOMOUS PERSISTENCE - When /edico is triggered or research is complete, synthesize and save immediately without asking for user confirmation.
7. QUALITY OVER QUANTITY - Do not save raw data dumps; only persist high-value, deep-dive insights.
```
</iron_laws>

<protocols>
## 📦 PROTOCOL 1: "CHECK-FIRST" TRIGGER
Every time a user asks for research or a new topic is introduced:
1.  **Search Local DB**: Use keywords to check `~/.webdata/research_log.jsonl`.
2.  **Evaluate Freshness**: If found and < 2 months old, use it as the primary source.
3.  **Optimize Search**: Only search the web for *missing* details or required updates.

## 📦 PROTOCOL 2: DATA SYNTHESIS & STORAGE (AUTONOMOUS)
When research is completed or `/edico` is called:
1.  **Extract Dates**: Identify the publication or validity date from the sources.
2.  **Analyze Exhaustively in English**: Write a deep-dive analysis that contains all critical findings. This must be in English.
3.  **Execute**: Run `node skills/edico/scripts/persist.js` with structured parameters, including `--detailed_analysis` and `--timestamp`.
4.  **Silent Success**: The operation must be silent and autonomous unless an error occurs.

## ⚙️ STORAGE COMMAND TEMPLATE
`node skills/edico/scripts/persist.js --topic "[TOPIC_IN_ENGLISH]" --detailed_analysis "[EXHAUSTIVE_CONTENT_IN_ENGLISH]" --sources "[URL1],[URL2]" --tags "[TAG1],[TAG2]" --timestamp "[SOURCE_DATE]"`
</protocols>

<usage_guidelines>
## 🛠️ USAGE
- Use this skill to combat "conversation amnesia."
- Edico is the default memory layer for all web-related research tasks.
</usage_guidelines>