NVIDIA/NemoClaw is an open-source reference stack for running supported AI agents inside NVIDIA OpenShell sandboxes with managed inference, network policies, integrations, snapshots, and lifecycle controls. It is used to run agents such as OpenClaw, Hermes, and LangChain Deep Agents with administrative and security controls. Catalogue add-ons guide coding agents through NemoClaw workflows.
Borrowing it
Nothing to install: this file belongs to NVIDIA/NemoClaw. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/NVIDIA/NemoClaw/main/.agents/skills/nemoclaw-contributor-create-pr/SKILL.mdgit clone --depth 1 https://github.com/NVIDIA/NemoClawWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nvidia/nemoclaw/nemoclaw-contributor-create-pr)<a href="https://agentmods.dev/skills/nvidia/nemoclaw/nemoclaw-contributor-create-pr"><img src="https://agentmods.dev/badge/skills/nvidia/nemoclaw/nemoclaw-contributor-create-pr/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nvidia/nemoclaw/nemoclaw-contributor-create-pr"><img src="https://agentmods.dev/badge/skills/nvidia/nemoclaw/nemoclaw-contributor-create-pr.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Snyk warn
- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00032 | $0.01329 |
| Opus 5.5 | $0.00013 | $0.00532 |
| Sonnet 5.5 | $0.00006 | $0.00266 |
| Haiku 4.5 | $0.00003 | $0.00133 |
Grade A, and why
nemoclaw-contributor-create-pr scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Create or Update a Pull Request
Session reporting follows v1 PR #13918 at 7238cbdff9, retaining its notices and report contract. Main retains its publication and validation requirements.
Before publication, follow Validation and Publication. For an open PR, first follow the PR follow-up contract; route valid code repairs to nemoclaw-contributor-implement-issue. These references own trusted-base validation, source-repository selection, guarded writes, verified commits, DCO, draft handling, and review follow-up. Follow Writing and review for PR text.
When changing live E2E evidence, require the implementation handoff to apply Define the Live Contract, or Move or Remove Evidence when pruning or relocating assertions. Route incomplete handoffs back to implementation.
-
Write the description from the canonical base’s
.github/PULL_REQUEST_TEMPLATE.md; keep the report out of it. Prepare a separate comment from.github/SESSION_REPORT_TEMPLATE.mdusing the evidence rules. Keep<details>withoutopen. -
Validate the comment:
rustc --edition=2024 tools/ci/pr-session-report.rs -o /tmp/pr-session-report /tmp/pr-session-report report-comment.md -
Publish through the publication procedure, preserving main’s description sections, DCO, and draft requirements. Read all pages of PR comments before posting, including after an uncertain retry; reuse an existing opening report by the PR author instead of duplicating it. If absent, post once as the PR author with
gh pr comment NUMBER --body-file report-comment.md. -
Read back the description and comments. Verify the description has no report, exactly one opening report by the PR author exists, and its contents match the intended report. Keep the exact summary
Authoring session — opening. Validate the retrieved files with/tmp/pr-session-report --pr description.md author-comment-1.md author-comment-2.md; pass every PR-author comment, including later reports and unrelated comments. Comments by others are ignored; a maintainer's comment cannot substitute for an automated PR author's report. -
Follow checks and automated reviews on the latest commit. If the metadata job fetched comments before the opening report was posted, wait for that run to finish, then use
gh run rerun --job JOB_IDfor its failedcommit-lintjob. Select the job from this PR's latest-revision run; a rerun retains the original revision but fetches current comments. Reruns require repository write permission, so external contributors may need a maintainer. Confirm the rerun passes.
CI fetches the description and all comment pages on PR opening, updates, reopening, and description/title edits. It checks the presence and format of the PR author's opening report, not the truth or completeness of its evidence. Fetch failures fail the check. Comment posting, editing, or deletion alone does not trigger a refresh; rerun the metadata job after such changes.
The check uses the candidate checkout's validator and workflow. It does not protect against a PR changing its own validation; trusted-base enforcement is outside this check's scope.
Preserve historical reports and their cutoffs. Optional later commit or merge reports may use the same collapsed format with a distinct summary and revision; they do not replace the opening report or create a publication requirement.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today Changed · +5 lines · +5 tokens per session ebe1cf760ab5
- 29d ago Changed · +3 lines ec76a0e7ec48
- 1mo ago Changed · +4 lines 1097b645c317
- 1mo ago Changed · +5 lines cf8afffa31aa
- 1mo ago Changed · -214 lines · -46 tokens per session 37c455057e6a
- 1mo ago Changed · +4 lines 4595ba806d78
- 1mo ago Changed · +100 lines d65fb8d18743
- 1mo ago Changed · +32 lines 6efb0d7004e2
- 2mo ago First seen · 130 lines · 73 tokens per session scan A fe09edae45d9
nemoclaw-contributor-create-pr is a skill published in the GitHub repository NVIDIA/NemoClaw (22,871 stars, last pushed today), licensed Apache-2.0. It adds 32 tokens to every session and 1,329 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
pentest-whitebox-code-review
Source code security audit using backward taint analysis, slot type classification, render context verification, and 3-phase parallel review producing an exploitation queue.
review-implement-phase
Implements triaged review actions, commits focused fixes, and posts Done plus resolves threads. Use when the user wants only the implementation phase of the review-framework workflow.
no-bare-casts
Writing as in TypeScript or TSX production code, modifying a file that contains a bare as cast, silencing a type error with a cast, encountering as unknown as, or reviewing a cast site.
update-pr
Update the pull request for the current session. Use when the user wants to push new changes to an existing PR.
performance-lint-rules
Performance review guidance for Oxc linter rule implementations. Use only when reviewing Rust rule code under crates/oxclinter/src/rules/ or when explicitly auditing those rules for performance improvements.
adversarial-reviewer
Stress-test a code change for concrete correctness defects, unsafe assumptions, and failure modes. Use when the user explicitly asks for an adversarial review, a deep bug hunt, or to tear a change apart, or to check your own change before submitting it. Do not use for ordinary review, style feedback, or general…