Skip to content
Back to skills

Jev Research Guard

ASecurity

Use `jev_check_research` after collecting evidence and before presenting a material claim as established.

  • 17 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 5, 2026
research

Security analysis

A100/100

Scanned October 5, 2026

npx -y skills add tensology/decisionsai --skill jev-research-guard --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Jev Research Guard?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Jev Research Guard
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tensology-jev-research-guard/badge)](https://www.skillsdirectory.com/skills/tensology-jev-research-guard)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
# Jev Research Guard

Use `jev_check_research` after collecting evidence and before presenting a material claim as established.

## Procedure

1. State exactly one claim to evaluate.
2. Provide compact evidence entries with source, date, and what each source directly supports.
3. Set `source_quality` conservatively. Use `unknown` when provenance is unclear.
4. Set `stakes` by the consequence of being wrong, not by confidence.
5. Call `jev_check_research`.

## Verdict handling

- `accept`: use the claim with appropriate attribution and caveats.
- `verify_more`: collect independent or more direct evidence, then evaluate again.
- `reject`: do not use the claim in its current form.

Do not send secrets, whole documents, or unrelated context. Jev evaluates supplied evidence; it does not browse or verify sources on its own.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…