fix(core): accept gpu: 0 in resources

[userAugustos]

Every canyonos deploy on main currently fails at config validation:

ERROR:canyonos_core:.car/config/global_controller.yaml:17: agents[0].resources.gpu: expected a finite number > 0, got the number 0
ERROR:canyonos_core:Configuration rejected: 7 problem(s) found; nothing was built.

#184 added the pre-deploy resource picker, which fills gpu: 0 for any agent that does not set it and writes the config back. #178 added the manifest schema, which requires gpu > 0. The two landed independently and contradict each other, so a config that does not already declare a positive gpu is rejected on every deploy. Behind the schema, cli._load_config has an older check with the same > 0 rule (Agent SplitAgent resource gpu must be a positive number.), so both are changed.

The runtimes already read a falsy gpu as "no GPU" (bool(resources.get("gpu")) in the global controller, if resources.get("gpu") before --gpus in the Local runtime), so gpu: 0 is a meaningful value, not a mistake. This makes both checks accept it: the gpu bound becomes inclusive (>= 0), while cpu and memory keep their exclusive > 0.

Repro: canyonos deploy on any project whose agents do not set resources.gpu (e.g. examples/portfolio), with the picker left at its defaults.

Verification: uv run --package canyonos-core pytest -q packages/core/tests/test_manifest_schema.py packages/core/tests/test_agent_declaration_schema.py packages/core/tests/test_cli.py — 139 passed.

Summary by CodeRabbit

• Bug Fixes
  • Resource settings now handle zero values consistently: GPU may be set to zero to indicate no GPU, while negative GPU values are rejected.
  • Validation continues to reject booleans, non-numeric values, and non-finite numbers for resource settings.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

Warning

Review limit reached

Next included review available in 35 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: 3baf918d-3fcf-404a-927c-0913f1e5342e

📥 Commits

Reviewing files that changed from the base of the PR and between eab6c04 and ce8a5d9.

⛔ Files ignored due to path filters (1)

• uv.lock is excluded by !**/*.lock

📒 Files selected for processing (6)

• packages/cli/canyonos/resources.py
• packages/cli/tests/test_resources.py
• packages/core/canyonos_core/cli.py
• packages/core/canyonos_core/schema/manifest.py
• packages/core/tests/test_cli.py
• packages/core/tests/test_manifest_schema.py

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: f5eb5791-f0be-40cf-b5af-7486d17b9540

📥 Commits

Reviewing files that changed from the base of the PR and between be91b2a and eab6c04.

📒 Files selected for processing (5)

• packages/core/canyonos_core/cli.py
• packages/core/canyonos_core/schema/_checks.py
• packages/core/canyonos_core/schema/manifest.py
• packages/core/tests/test_cli.py
• packages/core/tests/test_manifest_schema.py

Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

Resource validation now allows zero for GPU values and rejects negative GPU values. The numeric checker supports inclusive floors. Tests cover GPU zero and invalid resource values.

Changes

Resource validation

Layer / File(s)	Summary
Resource bounds and validation packages/core/canyonos_core/schema/_checks.py, packages/core/canyonos_core/schema/manifest.py, packages/core/canyonos_core/cli.py, packages/core/tests/test_manifest_schema.py, packages/core/tests/test_cli.py	The numeric checker supports inclusive floors. Manifest validation allows GPU zero while keeping CPU and memory strictly positive. CLI validation allows zero for CPU, memory, and GPU, and rejects negative values. Tests cover these resource bounds.

Priority: ➖ Normal

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix

Merge Risk: ⚪ Minimal · up to eab6c

The resource validation preserves positive CPU and memory requirements and permits GPU zero. No issue identified here needs to be resolved before merge.

Security Architecture Review

Security architecture risk: 🔵 Low · up to eab6c

Zero GPU now means no GPU across the checked validation and runtime paths. The change does not appear to grant GPU access or add a deployment entrypoint, but deployment access and production rollout conditions were not fully established.

Retained concerns
No architecture-level concerns identified.

Security review details

Security Blast Radius

• inferred — The checked effect is on agents deployed from a configured manifest. The newly accepted zero value does not request Local Docker GPU access; external authority to modify or deploy that manifest remains unverified.

Trust Boundaries and Controls

• observed — Manifest validation remains before build generation. The changed GPU bound permits zero but not negative values, and the runtime’s GPU option remains conditional on a truthy value.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 54.55% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 11 functions across 5 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the primary change: allowing gpu: 0 in core resource validation.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches 💡 1

📝 Generate docstrings 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[Saaketh0]

@userAugustos Basically if GPU is 0 it becomes None?

[userAugustos]

@Saaketh0 kinda (?)

before we had the validation with minimum that required gpu > 0 now with float it accepts gpu >= 0

This doesn't change runtime, because existing code already supported 0 gpu

global_controller.py: wants_gpu = bool(ctrl.get("resources", {}).get("gpu")) this would just be False

and then no --gpus

the only thing was that we (I) added on the manifest PR this check of bigger than 0 gpu, which is wrong

basically the whole code already supports "0 gpus" only the validation files changed

[Saaketh0]

great