Deep Researcher Release QA

What it does

When your agent uses it

• Validating an Deep Researcher Agent change before opening
• Merging a PR — choosing and running the right Python
• Eval checks for the surfaces you touched instead of one fixed command list

Example prompts

• “/deep-researcher-release-qa”

Requirements

• Python 3
• Compatibility (from SKILL.md): Claude Code, Codex, Cursor, OpenCode, and Agent Skills-compatible tools.
• Pre-approved tools (allowed-tools): Read, Bash

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

1. 1List the changed paths (git status, git diff --name-only) and map them to
2. 2Run the scoped Python, UI, docs, or eval checks for those surfaces.
3. 3If the change touches shared code or config, broaden to the full suite for
4. 4Re-run until lint, format, type, and tests all pass; fix failures rather than
5. 5Record the exact commands and their output for the PR description.

What it can do on your machine

Read from SKILL.md and the folder at commit 951a1a1. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • Read
  • Bash

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • uv
  • npm
  • git
  • ruff

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use uv, npm and git, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

• Compatibility

  Claude Code, Codex, Cursor, OpenCode, and Agent Skills-compatible tools.

  From compatibility in the SKILL.md frontmatter.

Context cost

Deep Researcher Release QA loads about 1.2k tokens when it runs, and up to ~2.1k if it reads all its reference files. Until then it costs about 61 tokens; SKILL.md has 488 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~61

When it runs · the whole SKILL.md, loaded when a task matches

~1.2k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~2.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• NoteMentions a .env fileSKILL.md:30

  - Never paste secrets or `deploy/.env` values into command output you share.

• NoteMentions a .env fileSKILL.md:111

  Forgetting that `nat eval` needs `deploy/.env`; run evals via

• NoteMentions a .env fileSKILL.md:112

  `dotenv -f deploy/.env run nat eval ...` (see the matrix reference).

• NotePre-approves every shell command (allowed-tools: Bash)SKILL.md

  allowed-tools: Read, Bash

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from NVIDIA-AI-Blueprints/deep-researcher-agent at commit 951a1a1, republished under its Apache-2.0 licence (© NVIDIA-AI-Blueprints). 488 words, ~1,164 tokens.

name

deep-researcher-release-qa

description

Use when validating an Deep Researcher Agent change before opening or merging a PR — choosing and running the right Python, frontend, docs, or eval checks for the surfaces you touched instead of one fixed command list.

allowed-tools

Read, Bash

compatibility

Claude Code, Codex, Cursor, OpenCode, and Agent Skills-compatible tools.

license

Apache-2.0

metadata.version

0.1.0

metadata.source-repo

NVIDIA-AI-Blueprints/deep-researcher-agent

metadata.tags

deep-researcher nemo-agent-toolkit validation testing release-qa

uv run ruff check <changed paths>          # lint
uv run ruff format --check <changed paths> # format check
uv run pytest <scoped test paths>          # tests

uv sync --project mcp --extra dev
uv run ruff check mcp
uv run ruff format --check mcp
uv run --project mcp --extra dev pytest mcp/tests

uv run ruff check .
uv run ruff format --check .
uv run pytest
uv run --project mcp --extra dev pytest mcp/tests

npm run lint
npm run type-check
npm run test:ci

© NVIDIA-AI-Blueprints, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (references) in .agents/skills/deep-researcher-release-qa of NVIDIA-AI-Blueprints/deep-researcher-agent.

• SKILL.md
• references/validation-matrix.md

Open the folder on GitHubat commit 951a1a1