Arcadia-1Arcadia-1/virtuoso-bridge-lite/AGENTS.md

virtuoso-bridge-lite

AGENTS.md1k starsChanged 9 days agoMITView on GitHub

Control Virtuoso through Python/SKILL and run Spectre independently. Spectre execution and PDK/CDF callbacks are the core. Lightweight schematic, symbol and GDS operations cover common work; raw SKILL handles the long tail. Existing one-host configuration and APIs remain sufficient. Multi-server, multi-account and process configuration are opt-in, with no mandatory migration. Before adding configuration or workflow abstractions, read the Lite decision. - Use uv and a project virtual environment. Install with uv pip install -e .. - Canonical implementation: src/virtuoso_bridge/; runnable workflows:…

Installs packages

Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository5 other files it gives its agents.
- netlistSkill
- optimizerSkill
- searchSkill
- spectreSkill
- virtuosoSkill

The file

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Distinctive:
Before adding configuration or workflow abstractions, read the Lite decision.
line 11
Very specific:
Use uv and a project virtual environment.
line 16
Very specific:
Use explicit schematic/layout create() or modify().
line 33
Distinctive:
Treat a shared CIW as user-owned.
line 37
Distinctive:
Keep authentication and daemon identity checks enabled on shared hosts.
line 42
Distinctive:
Run proportionate tests in the project environment:
line 68
Very specific:
Check pyproject.toml and .github/workflows/tests.yml for the current test configuration and platform coverage.
line 74

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

AGENTS.md

Agent guide — virtuoso-bridge-lite

Control Virtuoso through Python/SKILL and run Spectre independently.

Lite boundary

Spectre execution and PDK/CDF callbacks are the core. Lightweight schematic, symbol and GDS operations cover common work; raw SKILL handles the long tail. Existing one-host configuration and APIs remain sufficient. Multi-server, multi-account and process configuration are opt-in, with no mandatory migration. Before adding configuration or workflow abstractions, read the Lite decision.

Working in this repository

Use uv and a project virtual environment. Install with uv pip install -e ..
Canonical implementation: src/virtuoso_bridge/; runnable workflows: examples/; standalone helpers: tools/.
Locate modules with rg --files before guessing paths. CLI entrypoint: virtuoso_bridge.cli:main, registered as virtuoso-bridge in pyproject.toml. python -m virtuoso_bridge.cli is not the console entrypoint.
Preserve unrelated working-tree changes. Prefer an existing clean worktree for work that overlaps them.
Documentation edits and offline tests need no SSH, daemon startup or GUI interaction. Connect only when the task actually needs live EDA operations.
Installed skills may be symlinks to another checkout. Resolve their paths before editing; check the active client and daemon capabilities before using version-specific APIs. Updating instructions does not upgrade running code.

Operational boundaries

Execute SKILL through VirtuosoClient or virtuoso-bridge eval/load. Use explicit schematic/layout create() or modify(); replacing an existing design requires authorization. Check/save and refresh applicable PDK callbacks before netlisting; raw property changes alone do not reproduce GUI behavior.

Treat a shared CIW as user-owned. Inspect suspected dialogs out of band; let the user resolve unknown-origin dialogs. An uncertain request is not a failed request: preserve its handle/history and reconcile the original outcome before any retry. Read shared-CIW protection before popup-prone operations or recovery; its guard is opt-in and version-gated. Keep authentication and daemon identity checks enabled on shared hosts.

Read only the relevant references

Live Virtuoso/SKILL, schematic, symbol, layout, GDS or Maestro: virtuoso skill.
Standalone netlist-driven simulation and PSF: spectre skill. No Virtuoso daemon is required.
Netlist cleaning: netlist skill.
Parameter optimization: optimizer skill.
First connection, host/configuration changes or virtualenv rebuilding: connection guide.
SKILL functions, PDK parameters or version-dependent behavior: installed-doc verification.
Issues/PRs: use gh, then read tracker guidance. For issue triage also read label mapping.
Domain/interface design or review: domain decisions.
Explicit profile auditing or site environment selectors: profile audit; these are not default setup steps.
Requested traffic maintenance: traffic guide.
Windows skill links cloned as text: inspect scripts/fix-symlinks.sh before running it for the affected checkout. General user setup is in the guide.

Verification

Run proportionate tests in the project environment:

uv run --extra dev pytest

Check pyproject.toml and .github/workflows/tests.yml for the current test configuration and platform coverage. Shell/environment changes need real, bounded shell execution, not just command-string assertions; report skipped shell cases. CLI subprocess tests should exercise the registered entrypoint. Documentation-only changes need valid references and skill frontmatter checks.

State separately what was tested offline and what was verified on real Cadence/PDK hardware. A green fixture suite does not establish GUI-equivalent electrical results. Test layers, live-test variables and the rule against committing lab or PDK data are in tests/README.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

It worked
Partly
It didn't

What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free