data-ingestion — Claude Code skill: what it does, how to install
data-ingestion collects time-series points from CSV files or metric backends and converts them into standardized JSON.
Apache-2.0 ★ 151 updated 2026-10-12
/plugin marketplace add gemini-cli-extensions/sre
/plugin install sre-extension@sreWhat data-ingestion does
The skill accepts source_type and source_details input parameters. When handling CSV inputs, it verifies paths, generates intermediate output names in ~/.gemini/tmp/user/, and runs parse_csv.py on each file.
After processing individual inputs, it executes merge_timeseries.py to produce a single consolidated JSON structure, deleting temp files immediately after. Execution relies on Python shell scripts that need an active venv in the working environment.
What it helps with
- parse local CSV files containing time-series data
- merge multiple time-stamped metric logs into one JSON payload
- prepare telemetry traces for downstream evaluation skills
Example prompts for the agent
- Ingest time-series points from metrics_a.csv and metrics_b.csv
- Parse the CPU logs using data-ingestion with source_type: csv
- Run data-ingestion against access_metrics.csv and combine the output
How to install data-ingestion
1Plugin marketplace
/plugin marketplace add gemini-cli-extensions/sre/plugin install sre-extension@sreThis installs the whole sre-extension plugin with all its skills
2Skills CLI (Vercel)
npx skills add gemini-cli-extensions/sre --skill data-ingestion -a claude-code3Manually: put the files in place
~/.claude/skills/data-ingestion/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/data-ingestion/
1Skills CLI (Vercel)
npx skills add gemini-cli-extensions/sre --skill data-ingestion -a codex2Manually: put the files in place
.agents/skills/data-ingestion/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add gemini-cli-extensions/sre --skill data-ingestion -a cursor2Manually: put the files in place
.cursor/skills/data-ingestion/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install gemini-cli-extensions/sre skills/data-ingestion2Manually: put the files in place
.github/skills/data-ingestion/SKILL.md1Skills CLI (Vercel)
npx skills add gemini-cli-extensions/sre --skill data-ingestion -a gemini-cli2Platform CLI
gemini skills install https://github.com/gemini-cli-extensions/sre --path skills/data-ingestionManually: put the files in place
.devin/skills/data-ingestion/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about data-ingestion
How to install data-ingestion in Claude Code?
Run /plugin marketplace add gemini-cli-extensions/sre followed by /plugin install sre-extension@sre.
What does data-ingestion do?
It fetches time-series datasets from sources like CSV files and transforms them into unified JSON structures.
What environment is required to run the parser scripts?
A working Python venv is necessary because the skill invokes parse_csv.py and merge_timeseries.py directly.
Source text
SKILL.md · Apache-2.0
---
name: data-ingestion
description: 🐉 Fetches and parses time-series data from various sources.
metadata:
author: "Szymon Stawski + Gemini"
version: 0.2.0
status: draft
---
# Data Ingestion Skill
This skill is responsible for fetching and converting time-series data from different sources into a standardized JSON format for other skills to consume.
**Inputs:**
* `source_type`: String indicating the data source (e.g., "csv", "cloud_monitoring").
* `source_details`: A dictionary or list containing the necessary information to access the data source.
* For `source_type: "csv"`: A list of file paths.
**Output:**
A JSON string in the standardized format (see README.md for details).
**Workflow:**
1. **Validate Inputs:** Check if `source_type` and `source_details` are provided.
2. **Route to Parser:** Based on `source_type`:
* If `source_type` is "csv":
* Ensure `source_details` is a list of file paths.
* Create a list of temporary file names for intermediate JSON outputs (e.g., `~/.gemini/tmp/user/parsed_0.json`, ...).
* **Parse Each CSV:** Iterate through the input file paths:
* Execute `python ./skills/data-ingestion/parse_csv.py <input_csv_path> > <temp_json_path>` using `run_shell_command` (with venv activation).
* Check for errors.
* **Merge JSONs:** Execute `python ./skills/data-ingestion/merge_timeseries.py <temp_json_path_1> <temp_json_path_2> ...` using `run_shell_command` (with venv activation).
* Capture the stdout from `merge_timeseries.py` as the final result.
* **Clean up:** Remove the temporary JSON files.
* Handle any errors during script executions.
* If `source_type` is not supported, return an error message.
3. **Return JSON:** Output the standardized JSON string.
© Szymon Stawski + Gemini · Apache-2.0 · gemini-cli-extensions/sre