drift

Use after PR-sized work completes (or on /mek-drift) to compare auto scores against .mek/drift-baseline.json and surface divergences > 0.15.

Skill by mbachaud

IndexedApache-2.0★ 2

Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/mbachaud-drift

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

About

Use after PR-sized work completes (or on /mek-drift) to compare auto scores against .mek/drift-baseline.json and surface divergences > 0.15.

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools: No tools found
Secrets: None asked for
Install scripts: None found

Network hosts it talks to 1

codeload.github.com · found in install command

Files it touches

~/.claude/skills/drift

What's inside

Skill name: drift
Description: Use after PR-sized work completes (or on /mek-drift) to compare auto scores against .mek/drift-baseline.json and surface divergences > 0.15.
License: Apache-2.0
Instructions: 35 lines in SKILL.md · view on GitHub

Files (1)

SKILL.md

SKILL.md (first lines)

# drift — auto-vs-baseline divergence detection

## When to invoke

- Stop hook fires after a meaningful chunk of work (post_task_drift.py).
- User runs `/mek-drift`.

## What it does

1. Loads `.mek/drift-baseline.json` (created by `/mek-drift init`).
2. Runs the Python preset auto-scorers: test pass rate, lint, coverage. Security stays manual-only at v0.1.0.
3. Compares to manual baseline. Flags any dimension where:
   - `|auto - manual| > divergence_threshold` (default 0.15) AND `confidence >= min_confidence` (default 0.5), OR
   - `auto < floor` (ratchet violation).

## Resolution paths

- `/mek-drift accept` — agree with auto, update manual to match.
- `/mek-drift ack` — disagree with auto, dismiss this flag (record in baseline metadata).
- Re-audit manually and update the baseline.

## Config

`mek.toml > [drift]`:

```toml
[drift]
baseline_file = ".mek/drift-baseline.json"
preset = "python"
divergence_threshold = 0.15
min_confidence = 0.5
auto_run_on_stop = true
```

From mbachaud/MaxExpressKit, Apache-2.0.

Signals

Stars: 2
Forks: 0
Open issues: 0
Last commit (this folder): May 11, 2026 (5 months ago)
Last release: May 11, 2026 (5 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License: Apache-2.0
Repository: github.com/mbachaud/MaxExpressKit
Website: github.com/mbachaud/MaxExpressKit
First seen: Oct 24, 2026
Last checked: Oct 24, 2026
Source: GitHub github-graphql

More from mbachaud

Cite this page

drift skill. Agent Mart. https://agentmart.dev/skills/mbachaud-drift (last updated 2026-10-24)