Canonical: https://botbento.com/usecases/keep-coding-agent-guidance-matched-to-installed/
Format: Markdown representation of the public HTML page.

# Keep coding agent guidance matched to installed dependency versions

[Operations](https://botbento.com/usecases/ops/) · Intermediate

Check manifests and lockfiles against framework docs before the bot suggests code.

A bot reviews a repository's installed framework and library versions, compares them with the documentation and guidance an agent would normally follow, and flags places where behavior may differ. It maintains a short record of version-sensitive quirks and pinned test notes so future suggestions stay grounded in what is actually installed.

## Quick answers

**What does it do?** Check manifests and lockfiles against framework docs before the bot suggests code. Result: Version drift report.

**What does it need?** Tools: GitHub, Terminal, Docs. Intermediate: some setup. The prompt includes an approval step.

**How do I run it?** On a schedule: Weekly dependency check. Copy the prompt below into a bot in BotBento.

## How it works

1. On a schedule: Weekly dependency check
2. Then: Compare versions to docs
3. Result: Version drift report

## Prompt to start with

```text
Read the repository lockfiles and manifests to list installed framework and library versions. Compare each version with the guidance and documentation the agent would otherwise rely on. Flag mismatches where documented behavior differs from installed behavior. Record a known-defect pin or failing test note when behavior diverges. Produce a short drift report with affected files and suggested follow-up. Do not send messages, make payments, or delete files without explicit user approval.
```

## Tools it uses

GitHub, Terminal, Docs

## Works with

BotBento

## Tags

dependencies, versions, testing, documentation, drift

## Source

[Modern Web Guidance 1.0: The web platform expertise your coding agent is missing](https://news.ycombinator.com/item?id=50460398) — Hacker News · checked 2026-11-13

Summarized in our own words. Any results described there are the authors’ own, not measured by BotBento.

Verified 2026-11-13: passed every automated gate and a second-model review (kimi-code/k3.1-pro). Automated, not human review. [How verification works](https://botbento.com/usecases/verification/)

By BotBento Editorial · Added 2026-11-13 · Updated 2026-11-13 · [Suggest a correction](https://botbento.com/contact/)

## Related use cases

- [Independent verification for automated code fixes](https://botbento.com/usecases/independent-verification-for-automated-code-fixes/) — Validate coding agent patches against test suites before accepting modifications
- [Test developer docs by running the setup steps](https://botbento.com/usecases/automated-technical-documentation-stress-tester/) — Run each guide command in a sandbox and flag broken steps or missing context
- [Legacy code rewrite support with behavior mapping](https://botbento.com/usecases/legacy-code-rewrite-support-with-behavior-mapping/) — Map old behavior first, then draft refactors for engineers to review
