techmefr/mentis/data-pipeline-conventions
Claude Code and agent skill by techmefr/mentis
- community
Checking that the SKILL.md file is still in the repository.
Use when writing or reviewing a data pipeline (ETL/ELT), an analytical schema model, or a data quality validation. Distinct from transactional application code.
Install
npx vskill@latest install techmefr/mentis/data-pipeline-conventionsSKILL.md
The same SKILL.md was published earlier as techmefr/mentis/data-pipeline-conventions.
data-pipeline-conventions
Step 6 of the pipeline (WORKFLOW.md), for code that moves/transforms data between systems (ETL/ELT, analytical warehouse): distinct from the transactional application code (business CRUD) covered by the Laravel/NestJS conventions. Every rule below holds in a repo with nothing installed (CONVENTIONS.md, rule A).
An applied override is recorded, not hidden. Where a company's own data-engineering conventions or governance policy govern a rule here, write what they require, and leave one short neutral trace addressed to whoever can decide (a line in the PR description, the task summary or an ADR): which rule was overridden, by what, and that it is a resolved case. Never phrase it as "a conflict between mentis and the house rules" to a non-technical stakeholder — that framing reads as broken even when the case is a normal, resolved one, and has already caused a real project to get abandoned and restarted over nothing. Raise one specific, named question to the decider only when no rule anywhere actually resolves the case. What the landscape itself looks like before a query is safe to trust belongs to business/data-analytics.
When
As soon as a data pipeline, an analytical transformation, or a data schema definition meant for analysis is written or modified (not the app's own transactional database).
Steps
Read only the sections the task actually touches. The rules live one file per section under references/; §1 applies to every run and to any ad-hoc job on somebody's file, §2 is the validations, §3 is the shape of the tables, §4 is what it costs to run.
| § | Covers | Read it when | File | |
|---|---|---|---|---|
| 1 | Idempotence and reproducibility | any run, and any ad-hoc job on a supplied file | `01-idempotence.md` | |
| 2 | Data quality: verified, not assumed | writing validations, or facing a source you don't control | `02-data-quality.md` | |
| 3 | Analytical schema modelling | shaping tables meant for analysis | `03-modelling.md` | |
| 4 | Performance and cost | before a full reprocessing, or choosing a table's layout | `04-performance-cost.md` |
Output / checkpoint
Pipeline compliant with the four sections above; the data quality validations run and are green before the result is considered usable downstream, with the row counts in, out and rejected recorded (§2.7).
Guardrails
- Never write back over a supplied source file (§1.4), and never apply a destructive
transformation to someone's data without saying first what it changes and on how many rows (§1.5).
- Never run a destructive pipeline (full replacement of a production table) without explicit human
confirmation, and never treat a green sample as clearance for a full run (§1.12).
- Never let a validation failure pass silently (§2.1); where blocking is genuinely unacceptable,
quarantine the rows and publish the count rather than skipping them (§2.4).
- Never overwrite the raw layer (§3.1), and never lose a previous state with an in-place
UPDATE
where the history matters (§3.3).
- This block has no dedicated in-house production experience yet: to be confronted with the
first real data pipeline, not to be treated as proven doctrine.
Origin
Established dbt conventions, the DAMA-DMBOK quality dimensions and the classic SCD patterns, plus §1.4 and §1.5 from an org BI skill read for its handling discipline. The full provenance and the refresh log are in `references/origin.md`.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/data-pipeline-conventions/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/techmefr/mentis/data-pipeline-conventions)