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 bigbio/sdrf-skills --skill sdrf-annotate-reviewedgit clone --depth 1 https://github.com/bigbio/sdrf-skillsWrote 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/bigbio/sdrf-skills/sdrf-annotate-reviewed)<a href="https://agentmods.dev/skills/bigbio/sdrf-skills/sdrf-annotate-reviewed"><img src="https://agentmods.dev/badge/skills/bigbio/sdrf-skills/sdrf-annotate-reviewed/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/bigbio/sdrf-skills/sdrf-annotate-reviewed"><img src="https://agentmods.dev/badge/skills/bigbio/sdrf-skills/sdrf-annotate-reviewed.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- 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 | $0.00075 | $0.00667 |
| Opus 5 | $0.00037 | $0.00333 |
| Sonnet 5 | $0.00015 | $0.00133 |
| Haiku 4.5 | $0.00007 | $0.00067 |
Grade A, and why
sdrf-annotate-reviewed 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reviewed SDRF Annotation
Orchestrate annotation and independent review. Keep producer and reviewer contexts separate; validation by the producer is not review.
1. Produce the SDRF
Run the sdrf:annotate workflow in the current producer context. For an
existing artifact, run the applicable sdrf:fix, sdrf:review, or
sdrf:techrefine workflows first.
Create an evidence manifest using the schema in
../sdrf-adversarial-review/references/review-contract.md. Include exact source
URLs or local paths and map claims to SDRF columns or rows. Mark unavailable
evidence explicitly; do not invent citations.
2. Validate and mark pending
Run official template validation plus the repository's check and score tools. Fix deterministic errors before requesting review. Track the final artifact:
python3 <sdrf-skills-root>/tools/review_gate.py track <artifact> --cwd <repo-root>
3. Dispatch an isolated reviewer
Use a fresh subagent, hook agent, or equivalent isolated context. Pass only:
- the original request and acceptance criteria;
- the repository-relative SDRF path;
- the evidence-manifest path;
- the specification root and pinned revision;
- deterministic validation outputs; and
- the path to
sdrf-adversarial-review/SKILL.md.
Do not pass the producer transcript, reasoning, suspected issues, or proposed
verdict. Instruct the reviewer to use sdrf-adversarial-review, inspect raw
artifacts, and distrust producer assertions.
If the platform cannot create an isolated reviewer context, report that the adversarial gate is unavailable. Do not substitute producer self-review and do not claim a passing adversarial review.
4. Repair and re-review
When the reviewer returns blocker or important findings:
- Let the producer evaluate each finding against code, specification, and evidence; push back only with concrete counter-evidence.
- Apply accepted corrections.
- Re-run deterministic validation.
- Dispatch a new fresh reviewer. Never ask the original reviewer to rely on its previous verdict.
What ships with it
1 file 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.
- 11d ago First seen · 76 lines · 75 tokens per session scan A 0ef32896801f
sdrf-annotate-reviewed is a skill published in the GitHub repository bigbio/sdrf-skills (18 stars, last pushed 4d ago), licensed MIT. It adds 75 tokens to every session and 667 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
train-pose
Train/evaluate WiFi pose models honestly — camera-supervised (MediaPipe + CSI) and camera-free (WiFlow), always checked against the mean-pose baseline before any PCK is quoted.
openmetadata-pr-review
Reviews OpenMetadata and Collate pull requests as a maintainer deciding whether to merge. Use WHENEVER the user asks to review a pull request — any phrasing ("review this PR", "PR review", "can you review.
experiment-audit
A review step that checks whether an experiment's reported results are supported by real data and a sufficiently broad test.
test-enforcement
Use after implementing any feature or fix to ensure comprehensive test coverage. Enforces 90% line coverage in openmetadata-service, integration tests for all API endpoints in openmetadata-integration-tests, and Playwright E2E tests for UI changes.
playwright-validation
Use when validating UI changes in a branch require Playwright E2E testing. Reviews branch changes, validates UI with Playwright MCP, and adds missing test cases.