Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Tom409114/research-repo-doctor --skill rrdoctor-verifygit clone --depth 1 https://github.com/Tom409114/research-repo-doctorWrote 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/tom409114/research-repo-doctor/rrdoctor-verify)<a href="https://agentmods.dev/skills/tom409114/research-repo-doctor/rrdoctor-verify"><img src="https://agentmods.dev/badge/skills/tom409114/research-repo-doctor/rrdoctor-verify/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/tom409114/research-repo-doctor/rrdoctor-verify"><img src="https://agentmods.dev/badge/skills/tom409114/research-repo-doctor/rrdoctor-verify.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00028 | $0.00432 |
| Opus 5 | $0.00014 | $0.00216 |
| Sonnet 5 | $0.00006 | $0.00086 |
| Haiku 4.5 | $0.00003 | $0.00043 |
Grade A, and why
rrdoctor-verify 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 12d 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.
What it actually says
rrdoctor Verify Loop
Use this skill when a repository contains research code, notebooks, datasets, model artifacts, or an Artifact Evaluation / reproducibility checklist.
rrdoctor is the grader. The coding agent is the editor. Do not weaken rrdoctor checks to make the task pass.
Baseline
Run a deterministic baseline scan:
uvx rrdoctor scan . --format json --output baseline.json
If uvx is unavailable but rrdoctor is installed, use:
rrdoctor scan . --format json --output baseline.json
Plan
Generate a work order:
uvx rrdoctor plan . --output plan.md
Read plan.md and fix the highest-impact items first: missing run paths,
dependency metadata, local-only data paths, stale notebooks, citation/provenance
gaps, and randomness/seed issues.
Use rrdoctor fix --write only for safe scaffolding fixes. Generated files may
contain TODO placeholders that need human or agent completion.
Definition Of Done
After editing, run:
uvx rrdoctor scan . --baseline baseline.json --fail-on-new error
Stop only when this command passes. If it reports a new error, keep working from the finding and re-run the gate.
Trusted Dynamic Gate
The default scan is static: it does not install, import, build, or execute target repository code.
Only when the repository is trusted and the user wants dynamic evidence, run:
uvx rrdoctor verify . --run --timeout 600 --fail-on error
Treat this like running the repository yourself. Do not use it on arbitrary untrusted repositories.
Reporting Back
Summarize:
- the original readiness label and top findings,
- the files changed,
- the final baseline-gate result, and
- any remaining TODOs that require human judgment.
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.
- 12d ago First seen · 76 lines · 28 tokens per session scan A a117c1cc23d1
rrdoctor-verify is a skill published in the GitHub repository Tom409114/research-repo-doctor (1 stars, last pushed 2mo ago), licensed MIT. It adds 28 tokens to every session and 432 once invoked, about $0.0001 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.
Other skills, from other repositories
conventions
Calkit conventions and foundational context. Load whenever working in a Calkit project—defines calkit.yaml structure, environments, pipeline stages, CLI commands, and version control conventions.
build-paper-pipeline
Build the pipeline stages a manuscript's TODO comments ask for.
check-reproducibility
Check whether a project is fully traceable, and fix what isn't, including numbers typed into a manuscript that no pipeline output accounts for. Use when the user invokes /calkit:check-reproducibility, asks whether a project is reproducible, or asks where a number in a paper came from.
create-pipeline
Convert an existing repo with ad hoc scripts into a fully reproducible Calkit pipeline. Use when the user invokes /calkit:create-pipeline or asks to make a project reproducible end-to-end.
add-pipeline-stage
Add a single new stage to an existing Calkit pipeline. Use when the user invokes /calkit:add-pipeline-stage or asks to add a script, notebook, or command to the pipeline.
check-questions
Review a Calkit project's questions and answers against their evidence. Use when the user invokes /calkit:check-questions, asks whether the project's answers are still true, or after a pipeline run changes results that answers cite.