Skip to content
Skills Library
Data & analytics · Learning & explanation

using-wraith

/using-wraith

Use when the user asks what wraith does or which wraith skill or agent fits a data or GCP task. Lists the wraith skills and agents and the rules they share.

Licence
MIT
Last repository change
13.11.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 13.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when the user asks what wraith does or which wraith skill or agent fits a data or GCP task. Lists the wraith skills and agents and the rules they share.

Original author description

Use when the user asks what wraith does or which wraith skill or agent fits a data or GCP task. Lists the wraith skills and agents and the rules they share.

How it works

  1. Instruction priority
  2. Shared rules
  3. Skills
  4. Agents
  5. Companions
  6. Safety
Original SKILL.md headings · 60 lines, 439 words
SKILL.md preview
---
name: using-wraith
description: "Use when the user asks what wraith does or which wraith skill or agent fits a data or GCP task. Lists the wraith skills and agents and the rules they share."
license: MIT
metadata:
  source: GS@1bd2330:ETHOS.md
---

# Using wraith

wraith is a plugin of skills and agents for data and infrastructure work on GCP: dbt on BigQuery, ODCS data contracts, ontologies, Cloud Run and Terraform.

## Instruction priority

The user's direct requests first, then the project's AGENTS.md and CLAUDE.md, then wraith skills, then companion skills, then defaults.

## Shared rules

- Project specifics come from the `## wraith` block in the project's AGENTS.md. Never guess a command, project, region or target; if the block is missing, ask once and suggest wraith:setup.
- AI recommends, the user decides: production changes, pushes, deletes and spending need the user's explicit go-ahead in this conversation.
- A number without its mechanism is a rumour: say how every number was measured.
- Anything that touches state outside the process will one day run twice, so make it idempotent.
- Before destructive data and infra commands, follow wraith:careful.

## Skills

| Need | Skill |
|---|---|
| Review a plan that touches data | wraith:plan-data-review |
| Review a plan that touches GCP infrastructure | wraith:plan-infra-review |
| Write or change a data contract | wraith:odcs-contracts |
| Build or fix dbt models on BigQuery | wraith:dbt-bigquery |
| BigQuery cost, access or erasure | wraith:bigquery |
| Cloud Run services and jobs | wraith:cloud-run |
| Terraform on GCP | wraith:terraform-gcp |
| What business entities mean | wraith:ontology-modeling |
| Make the warehouse usable by AI agents | wraith:context-layer |
| IAM, secrets, ingress or personal data | wraith:security-audit |
| Prove a change against a dev target | wraith:qa |
| Production rollout | wraith:deploy |
| Destructive or production commands | wraith:careful |
| Configure wraith for a repository | wraith:setup |
| This list | wraith:using-wraith |

## Agents

| Agent | Dispatch when |
|---|---|
| wraith:data-engineer | dbt on BigQuery, ODCS contracts, pipelines, backfills and data quality. |
| wraith:platform-engineer | Terraform, Cloud Run, IAM, Workload Identity Federation and CI/CD. |
| wraith:sre | Observability, alerts, SLOs, incident triage and post-deploy watches. |
| wraith:security-engineer | Security and privacy review of GCP and data work. |
| wraith:ontologist | Ontology and context-layer work. |

## Companions

wraith names companion skills and never copies them. All are optional; every wraith skill works without them.

- Vendor mechanics: dbt Labs, Google and HashiCorp skills, listed in each wraith skill.
- General workflow (designs, plans, test-first work, debugging, code review, handoffs): obra/superpowers and mattpocock/skills.

Where a companion and a wraith skill disagree on data or GCP work, the wraith rule wins.

## Safety

wraith:careful is an instruction, not a security boundary. Keep the host's permission prompts on.

Versions

The database has no other skills named using-wraith.

Database snapshot: 13.11.2026 · package version 135f07317710 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100setup
    Use once per repository before relying on the other wraith skills, or when the user asks to configure wraith. Records the project's commands, GCP projects, dbt targets and conventions in an AGENTS.md wraith block that every wraith skill reads, and recommends vendor companions.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.