Dr-Claude-Code: Command for Claude Code

.claude/commands/raca/harvest-and-report.md

harvest-and-report is a command for Claude Code from Zayne-sprague/Dr-Claude-Code. It costs 31 tokens per session (798 once invoked), scanned A, original, MIT.

A post-run workflow for collecting experiment outputs, checking them, uploading them to Hugging Face, synchronizing a dashboard, and notifying the user. Hugging Face is a platform for sharing machine-learning models and datasets.

In plain words
What is it for?
Use it whenever an experiment creates artifacts to download or locate results, validate samples and scientific sense, upload them, update the dashboard, and report anomalies.
Why use it?
It prevents useful results from being left on a cluster or published without checking their contents, and it allows partial results to be handled before a job ends.

Command for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions subagents.

This is Zayne-sprague/Dr-Claude-Code's own configuration. It tells Claude Code how to work on Dr-Claude-Code itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Dr-Claude-Code configures →

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is raca download <cluster> <working_dir>/results/ ./local_results/<experiment>/.

Reuse

Borrowing it

Nothing to install: this file belongs to Zayne-sprague/Dr-Claude-Code. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/Zayne-sprague/Dr-Claude-Code/main/.claude/commands/raca/harvest-and-report.md
Clone the repo
git clone --depth 1 https://github.com/Zayne-sprague/Dr-Claude-Code

Made for: Claude Code.

Wrote 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.

agentmods badge for harvest-and-report

README.md
[![agentmods](https://agentmods.dev/badge/commands/zayne-sprague/dr-claude-code/harvest-and-report/github.svg)](https://agentmods.dev/commands/zayne-sprague/dr-claude-code/harvest-and-report)
Your own site
<a href="https://agentmods.dev/commands/zayne-sprague/dr-claude-code/harvest-and-report"><img src="https://agentmods.dev/badge/commands/zayne-sprague/dr-claude-code/harvest-and-report/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.

agentmods 80×15 button for harvest-and-report

Your own site · 80×15
<a href="https://agentmods.dev/commands/zayne-sprague/dr-claude-code/harvest-and-report"><img src="https://agentmods.dev/badge/commands/zayne-sprague/dr-claude-code/harvest-and-report.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 798 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00031 $0.00798
Opus 5 $0.00015 $0.00399
Sonnet 5 $0.00006 $0.00160
Haiku 4.5 $0.00003 $0.00080

Measured 9d ago against content hash 3b1504fb37ea, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

harvest-and-report 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 9d ago.

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.

.claude/commands/raca/harvest-and-report.md · 89 lines

How it starts

The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Harvest & Report

Run this whenever an experiment produces artifacts — partial results during a job, final results after completion, or anything in between. Don't wait for the job to finish.

This is a FLEXIBLE workflow — adapt to the experiment type, but never skip validation or dashboard sync.

Step 1: Get the artifacts

If the job is on a cluster:

raca ssh <cluster> "ls <working_dir>/results/"
raca download <cluster> <working_dir>/results/ ./local_results/<experiment>/

If artifacts are already local, just locate them.

Step 2: Validate

Read the Red Team Brief at notes/experiments/<experiment>/red_team_brief.md.

Dispatch a data-validator subagent:

  • Pass the validation criteria from the brief
  • Sample 20-50 rows
  • Check for: truncation, degenerate repetition, suspicious scores, format violations, missing fields

You also review: does the data make scientific sense? Not just format — substance. Compare against what the experiment was supposed to produce.

Anomalies don't block the harvest — they get flagged. But critical issues (all outputs truncated, wrong model loaded, scores nonsensical) should be raised to the user immediately.

Step 3: Upload to HuggingFace

from hf_utility import push_dataset_to_hub

push_dataset_to_hub(
    dataset=dataset,
    dataset_name="<experiment-slug>-<description>-<version>",
    experiment_slug="<experiment-slug>",  # must match experiment folder name
    metadata={
        "script_name": "<the script that generated this>",
        "model": "<model used>",
        "description": "<what this dataset contains — note if partial>",
        "experiment_name": "<experiment-slug>",
        "job_id": "<cluster:job_id>",
        "cluster": "<cluster>",
        "artifact_status": "partial",  # or "final"
        "canary": False,
    },
    tags=["<experiment-name>", "<condition>"],
    column_descriptions={<column: description for each column>},
)

Follow .claude/rules/huggingface.md.

For partial results during a running job: append to the existing HF dataset rather than creating a new repo each time. But still alert the user that new rows are available.

Read the full file on GitHub · 89 lines

Changes

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.

  1. 9d ago First seen · 89 lines · 31 tokens per session scan A 3b1504fb37ea

Subscribe to this mod's changes

harvest-and-report is a command published in the GitHub repository Zayne-sprague/Dr-Claude-Code (5 stars, last pushed 5mo ago), licensed MIT. It adds 31 tokens to every session and 798 once invoked, about $0.0002 per session on Opus 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 2026-08-31.