bodies; open a pull request or link a branch when code is available.
- Before opening a pull request, review `CONTRIBUTING.md`, follow the PR template, keep the change focused, and summarize
tests to verify functionality
- **Error Handling**: Handle errors gracefully with proper telemetry logging
- **Review Process**: Code must be reviewed by at least one other developer before merging
### Omnichannel Chat
benchmarks/` (`make
benchmark`, `make profile`, `make benchmark-save` / `make benchmark-compare`
for regression checks — review `DEVELOPMENT.md`). When touching content rules,
the lint tree, or `utils.py` read paths, save a baseline
pool.acquire()` — never hold `Arc ` across await points.
### Review Checklist for DB Changes
When reviewingcode that touches ChronDB:
- [ ] Key prefix is unique and documented in the table above
- [ ] Serialization round
Repository Instructions
Follow `AGENTS.md` for project-wide coding, testing, documentation, changelog,
and commit rules.
For release preparation, publishing, verification, or recovery tasks, read
`.agents/skills/project-release/SKILL.md` before making changes. That file
sentence rationale. Keep the `(#NNN)` reference; fix any sibling entry the change makes stale.
### CodeReview Mindset
- Be cautious about deleting code; avoid `git restore` without review
- Do not create
applyTo: "**"
---
# Copilot CodeReview Instructions
## Project Context
Zeph is a Rust AI agent workspace (Edition 2024, resolver 3) with 31 library crates under `crates/` plus the `zeph` CLI binary. Review
style and help contributors produce high-quality, consistent code. Always consider existing patterns in the repository—when in doubt, review similar resources or tests for reference and keep
viberaven clean --plan` to create a reviewable cleanup plan.
- Provider dashboard checks are not cleared by repo-code edits. Billing/product configuration, DNS, webhooks, credentials, quotas, and live provider verification must
working on a pull request, **DO NOT** mark the PR as ready for review until all of the following conditions are met:
1. **All CI/CD pipelines must pass**: Wait
check your capability profile in `.ai-team/team.md` under the **Coding Agent → Capabilities** section.
- **🟢 Good fit** — proceed autonomously.
- **🟡 Needs review** — proceed, but note in the PR description that a squad
setup, but use different prompts, expected outputs, and evaluation pipelines. Future categories such as `code-review` can be added within the same overall benchmark structure, though they may require different
Copilot codereview instructions
You review pull requests for this repository. `AGENTS.md` at the repository
root, the package `AGENTS.md` files, and the Serena memories under
`.serena/memories/` are the coding standard
wrapper is shell + JS, published to npm.
Any Copilot agent (Coding Agent, CodeReview, ci-doctor) working on this repo must follow the rules below.
## Hard rules — never violate
with:
```bash
./setup-hooks.sh
```
## Code Quality — No AI Slop
- No verbose comments restating what the code does — Rust is expressive enough
- No filler phrases: "This ensures that...", "In order
rename variables, restructure working code, or refactor beyond the task at hand. Keep diffs small and focused so they are easy to review.
### Kotlin/Ktor Patterns
- ApplicationBuilder pattern for bootstrapping
- Sealed