How do I install the Conversation Flow skill?

Run npx skillmds@latest add reutfarkash/conversation-flow in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Conversation Flow safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Conversation Flow?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Conversation Flow free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Conversation Flow?

ReutFarkash (@reutfarkash) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Conversation Flow by reutfarkash — Agent Skill

Name: Conversation Flow
Rating: 4.2 (5 reviews)
Author: ReutFarkash

File contents

Conversation Flow (v3.3 - Architect Refined)

Overview

This skill generates a high-fidelity narrative map that prioritizes Resolution Persistence and Visual Clarity. It uses a structured topological approach to link technical turns, project shifts, and strategic decisions.

Workflow: Generating a v3.3 Flow Analysis

Step 1: Gathering & Partitioning

Identify Projects: Analyze the session to group turns by project (e.g., Skills Dev, Retrospective Recovery).
Context Warm-up:
- Read ../_shared-gemini/skill_settings.md for templates.
- Read meta/SESSION_FLIGHT_RECORDER.md for Git SHAs and outcomes.

Step 2: Visual Taxonomy & Styling

Apply the visual standards and Mermaid classDef blocks defined in mermaid_standards.md.

MANDATORY: Group related nodes into subgraph ProjectName.
MANDATORY: Use internal Obsidian links (<a class='internal-link'>) for session persistence.

Step 3: Syntax Generation

Generate the Mermaid graph ensuring all labels are double-quoted and backticks are triple-escaped.
List active skills in the [!ABSTRACT] callout or a floating subgraph. Do NOT link skill nodes to the narrative flow.

Step 4: Pre-Flight Validation & Correction Loop

Run Internal Validation: After drafting the output file but BEFORE finishing the session, you MUST: node scripts/validate_mermaid.cjs <path_to_draft_file>
Internal Correction: If the validator outputs any errors (e.g., "Label must be wrapped in double quotes"):
- You MUST perform the fix internally using the replace tool.
- DO NOT report the errors to the user as a final failure; treat this as an internal development step.
Final Verification: Only present the finalized, error-free output to the user once the validation script returns: ✅ Mermaid syntax is valid.

Step 5: Final Delivery

Provide the finalized, verified [!ABSTRACT] callout and Mermaid map.
Provide the chat save tag as the final, unstyled line of the response: #chat/save/YYYY-MM-DD_HHMMSS-description.

Output Structure

YAML Frontmatter
[!ABSTRACT] Callout (Metadata, Active Skills, versions)
Mermaid Topological Map (with subgraphs)
Textual Breakdown (Summary, Open Loops, Decisions)

---name: conversation-flowdescription: Use when the user needs to visualize a complex session, map narrative pivots, or identify open loops and technical decisions. Provides a high-resolution, HTML-linked topological map of the conversation.---# Conversation Flow (v3.3 - Architect Refined)## OverviewThis skill generates a high-fidelity narrative map that prioritizes **Resolution Persistence** and **Visual Clarity**. It uses a structured topological approach to link technical turns, project shifts, and strategic decisions.## Workflow: Generating a v3.3 Flow Analysis### Step 1: Gathering & Partitioning1. **Identify Projects:** Analyze the session to group turns by project (e.g., `Skills Dev`, `Retrospective Recovery`).2. **Context Warm-up:** - Read `../_shared-gemini/skill_settings.md` for templates. - Read `meta/SESSION_FLIGHT_RECORDER.md` for Git SHAs and outcomes.### Step 2: Visual Taxonomy & StylingApply the visual standards and Mermaid `classDef` blocks defined in **[mermaid_standards.md](references/mermaid_standards.md)**.- **MANDATORY:** Group related nodes into `subgraph ProjectName`.- **MANDATORY:** Use internal Obsidian links (`<a class='internal-link'>`) for session persistence.### Step 3: Syntax Generation1. Generate the Mermaid graph ensuring all labels are double-quoted and backticks are triple-escaped.2. List active skills in the `[!ABSTRACT]` callout or a floating subgraph. Do NOT link skill nodes to the narrative flow.### Step 4: Pre-Flight Validation & Correction Loop1. **Run Internal Validation:** After drafting the output file but BEFORE finishing the session, you MUST: `node scripts/validate_mermaid.cjs <path_to_draft_file>`2. **Internal Correction:** If the validator outputs any errors (e.g., "Label must be wrapped in double quotes"): - You MUST perform the fix internally using the `replace` tool. - DO NOT report the errors to the user as a final failure; treat this as an internal development step.3. **Final Verification:** Only present the finalized, error-free output to the user once the validation script returns: `✅ Mermaid syntax is valid.`### Step 5: Final Delivery1. Provide the finalized, verified `[!ABSTRACT]` callout and Mermaid map.2. Provide the chat save tag as the final, unstyled line of the response: `#chat/save/YYYY-MM-DD_HHMMSS-description`.## Output Structure1. YAML Frontmatter2. [!ABSTRACT] Callout (Metadata, Active Skills, versions)3. Mermaid Topological Map (with subgraphs)4. Textual Breakdown (Summary, Open Loops, Decisions)

Frequently asked questions

Conversation Flow