claude

agent-write-guard

Claude Code only. Optional. Denies the implementer agent file edits on the test tree and the test-author agent file edits on production code. Without it, both agents are bounded by their instructions only.

3 on GitHubby Victor Macaubasupdated 9 days ago
/

Plugin agent-write-guard from marketplace data-engineering-skills in victormacaubas/data-engineering-skills. Commands read from the repo's manifest — 8 days ago.

What it installs

No skills, commands, agents, hooks or MCP servers were found in the standard plugin folders.

View source on GitHub

More from victormacaubas/data-engineering-skills

All 5 plugins →

flow

claude

How work gets driven from an unformed idea to something tracked and built: exploring, pressure-testing, delegating implementation, ticketing, and capture. Invoke as /flow:<skill-name>. The scout skill requires the separately installed pathfinder and researcher agents; orchestrate requires implementer, pathfinder, researcher, structure-reviewer, and test-author.

6 skills

3

craft

claude

How code gets written and checked: architectural baselines, Python standards, and the defect and structure reviews that gate a finished change. Invoke as /craft:<skill-name>. The architecture-baseline skill requires the separately installed researcher agent.

4 skills

3

data

claude

The warehouse surface: canonical SQL standards for trustworthy analytics, and Snowflake ACCOUNT_USAGE auditing for masking, row access, classification, tags, grants, logins, and query history. Invoke as /data:<skill-name>.

2 skills

3

Claude Code only. Shows the SQL that the main session and its subagents run through the snowflake-raw MCP server in a pane, grouped by agent. Requires the snowflake-raw MCP server.

3

Maintain agent-write-guard?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins5 plugins