feat : add optimize-simplicite-log skill (#1742)

skills/optimize-simplicite-logs/SKILL.md

@@ -0,0 +1,74 @@
+---
+name: optimize-simplicite-logs
+description: capability to parse Simplicité logs from a raw `.txt` file, filter fields to reduce noise, and output the result as structured JSON.
+---
+
+# Optimize Simplicite Logs
+
+This skill provides the capability to parse Simplicité logs from a raw `.txt` file, filter fields to reduce noise, and output the result as structured JSON. This is critical for optimizing AI context size (saving ~56% of tokens) and providing structured, predictable data for troubleshooting.
+
+## When to Use This Skill
+
+Use this skill when you need to:
+- Analyze user-provided Simplicité log files in `.txt` format.
+- Avoid ingesting massive raw log files into your context window.
+- Extract structured fields (like `timestamp`, `level`, `body`) from verbose multi-line log output.
+
+**IMPORTANT:** Instead of directly reading a raw `.txt` log file provided by the user using file read tools, you **must** use one of the log converter scripts (PowerShell or Python) to parse the file into a JSON format first, optionally extracting only the fields needed.
+
+## Prerequisites
+
+- Access to either the PowerShell script (`/scripts/SimpliciteLog2Json.ps1`) or the Python script (`/scripts/simplicite-log2json.py`).
+
+## Core Capabilities
+
+### 1. Context Optimization
+Reduces the tokens consumed by large Simplicité logs by extracting only relevant log fields (e.g. `body`, `timestamp`, `level`) and discarding non-relevant structural log data (like `app`, `endpoint`, `contextPath`).
+
+### 2. Multi-line Support
+Properly captures stack traces and multiline errors inside the `body` field of the JSON structure, which a simple text search might miss.
+
+### 3. Stdout Support
+If no output path is provided for the JSON file (e.g. omitting `--output` or `-Output`), the parsed JSON will be printed directly to stdout, allowing you to pipe the output to other tools.
+
+## Output Summary
+
+After processing, the tool prints a summary to stderr (or console):
+```
+Processed: 123 entries, Skipped: 2 entries
+```
+
+## Usage Examples
+
+### Example 1: Python Version (Recommended)
+Convert a log file to JSON, keeping only the most important fields:
+```sh
+python /absolute/path/to/skills/optimize-simplicite-logs/scripts/simplicite-log2json.py <input.txt> --include timestamp,level,body --output <output.json>
+```
+
+### Example 2: PowerShell Version
+```powershell
+/python /absolute/path/to/skills/optimize-simplicite-logs/scripts/SimpliciteLog2Json.ps1 -InputPath "<input.txt>" -Output "<output.json>" -Include "body,timestamp,level"
+```
+
+After generating the `<output.json>`, you can safely read the resulting file to perform your analysis.
+
+## Guidelines
+
+1. **Always Convert First:** Never directly read `.txt` log files from Simplicité using standard text reading tools. Always convert them to JSON using the available scripts.
+2. **Filter Fields:** Use `--include` (Python) or `-Include` (PowerShell) to restrict fields to what is absolutely necessary to diagnose the issue (usually `timestamp,level,body`).
+3. **Available Fields:** The fields you can filter include: `timestamp`, `app`, `level`, `endpoint`, `contextPath`, `event`, `user`, `class`, `function`, `rowId`, `body`.
+
+## Common Patterns
+
+### Pattern: Fast Contextual Troubleshooting
+```sh
+# 1. Run the script to generate a minified JSON output in the current directory
+python /absolute/path/to/skills/optimize-simplicite-logs/scripts/simplicite-log2json.py logs.txt --include timestamp,level,body --output logs_minified.json
+
+# 2. Then read logs_minified.json to understand the context.
+```
+
+## Limitations
+
+- The parser depends on a fixed regex pattern that matches the standard Simplicité log output. If the log format has been heavily customized, parsing might fail or degrade.