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 magnus919/agent-skills --skill training-data-annotationgit clone --depth 1 https://github.com/magnus919/agent-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/magnus919/agent-skills/training-data-annotation)<a href="https://agentmods.dev/skills/magnus919/agent-skills/training-data-annotation"><img src="https://agentmods.dev/badge/skills/magnus919/agent-skills/training-data-annotation/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/magnus919/agent-skills/training-data-annotation"><img src="https://agentmods.dev/badge/skills/magnus919/agent-skills/training-data-annotation.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 | $0.00093 | $0.01105 |
| Opus 5.5 | $0.00037 | $0.00442 |
| Sonnet 5.5 | $0.00019 | $0.00221 |
| Haiku 4.5 | $0.00009 | $0.00111 |
Grade A, and why
training-data-annotation 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 1mo 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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Training Data Annotation
Treat annotation as a measurement system and a human workflow. The objective is useful, auditable labels for a defined downstream decision, not the largest possible label count.
Workflow
- Define the downstream task, population, label unit, label set, edge cases, abstention label, and what constitutes an unusable item. Separate training, validation, held-out evaluation, and challenge data before sampling.
- Write an annotation guide with positive and negative examples, boundary rules, source requirements, escalation, and a change log. Have domain owners approve consequential definitions.
- Establish provenance and rights: source, collection time, license/consent, transformations, sensitive fields, access, retention/deletion, and dataset version. Do not move data into a labeling system until this record is clear.
- Acquire items using a declared mix of representative sampling and targeted acquisition. Uncertainty finds confusing items; diversity and novelty find gaps; stratification protects rare or high-risk slices. Keep acquisition batches identifiable and retain a representative held-out audit sample.
- Qualify annotators for the task, train them on calibration items, and measure quality and workload. Record role and relevant expertise in restricted metadata. Do not use agreement as a universal truth score.
- Run blinded calibration and periodic blind rechecks. Compare labels with adjudicated references where available; preserve meaningful disagreement, ambiguity, and abstention instead of forcing consensus.
- Review disagreement by item, label, slice, annotator, and guide version. Adjudicate only with a stated rule and qualified authority. Feed confirmed guide defects back into training and re-review affected labels.
- Stop or continue based on marginal downstream value, coverage, quality, cost, and time. An item quota may stop spending but cannot by itself establish sufficient quality. Freeze the manifest and hand off to data-engineering for storage/orchestration and ml-engineering for training use.
What ships with it
10 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.
- evals/evals.json 3.5 KB
- README.md 1.6 KB
- references/acquisition-and-stopping.md 951 B
- references/annotator-quality.md 1.0 KB
- references/provenance-and-rights.md 688 B
- references/source-index.md 1.0 KB
- references/task-and-label-design.md 898 B
- templates/annotation-guide.md 383 B
- templates/annotation-plan.md 568 B
- templates/quality-report.md 548 B
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.
- 1mo ago First seen · 99 lines · 93 tokens per session scan A c1fa2d0ab64f
training-data-annotation is a skill published in the GitHub repository magnus919/agent-skills (133 stars, last pushed 2d ago), licensed MIT. It adds 93 tokens to every session and 1,105 once invoked, about $0.0004 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 2026-09-15.
Other skills, from other repositories
map
Build and commit a Cortex function knowledge graph — maps structural dependencies and domain intent relationships across all AI functions in the project. Supports --reduce (default on) for transitive reduction of the dependency graph.
context-add
Guided wizard to capture or update organizational Snowflake/Cortex standards into .cocoplus/context/ .md. Presents a menu of 6 standard types, collects answers via multi-turn dialogue, and commits the file.
cocolean
CocoLean — minimum viable Cortex surface discipline. Pre-build decision ladder and intensity mode management. Invoke with $lean, $lean lite, $lean full, or $lean ultra.
test
Enter the Test phase of CocoBrew. Reads spec.md test requirements, generates test cases, executes SQL validation and quality checks, records results in test.md. Can be re-run without full rebuild. Requires Build phase completion.
map-diff
Analyze the impact of staged git changes against the committed Cortex function knowledge graph — shows which downstream functions are affected before you commit.
map-explain
Produce a natural-language explanation of a specific Cortex function, business capability, or schema element from the committed knowledge graph.