data-report Eval report.

Validation plan

What the agent should do next

1. 1Inspect repository, README/SKILL.md, license, and recent commits before production use.
2. 2Install in an isolated workspace or sandbox with no production secrets available.
3. 3Run the smallest representative task and record files touched, commands run, network access, and outputs.
4. 4Compare the selected skill against at least one alternative when the eval status is review or failed.
5. 5Promote only after the agent reports a successful verification result and unresolved warnings are accepted.

Do not use when

Conditions that require another skill

• teams that need a vendor-supported SLA
• production agents without a repository review
• No explicit sanitization/escaping rule: user data is inserted into an HTML/JS report, so malicious CSV/Excel/JSON strings could become HTML or script if not escaped and JSON-encoded.
• The tracked source changed or could not be synchronized. Review the current source before installing.
• Parsing workflow is underspecified: SKILL.md says to parse user input but does not explain how to handle CSV delimiters, Excel conversion, JSON schema normalization, missing values, or date formats.
• Limitations are missing: no guidance for large datasets, offline/no-CDN environments, non-numeric data, or cases where year-over-year comparison is unavailable despite the template requiring 同比变化.