Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →florianbruniaux/update-threat-db
Research and update the AI agent security threat intelligence database
Passed 20 of 20 test cases with the skill, and 6 of 20 without it.
and 14 more in the full report
14 fixed6 passed either way0 still failing0 broke3 errored
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
What specific JSON file artifact should be synchronized from AgentSec to the guide and landing repositories once local tests pass, and what condition must be met before copying it?
| Checked | Without skill | With skill |
|---|---|---|
| Identifies exports/security-feed.v1.json as the file to synchronize. | Missed | Passed |
| Requires AgentSec local tests to pass prior to synchronization. | Passed | Passed |
The instructions your model reads when the skill loads.
AgentSec Triage owns the technical source of truth. This guide skill is a delegator and does not carry a private copy of the threat database.
agentsec-triage checkout or use AGENTSEC_REPO.AGENTS.md and
.claude/commands/update-threat-db.md completely.exports/security-feed.v1.json to the guide and landing only
after AgentSec passes locally..claude/commands/update-threat-db.md for compatibility
database updates and the landing's searchable catalogue sync/check command.
A current feed badge does not prove the catalogue contains the new records.Do not treat the guide's compatibility database as canonical. Do not publish, tag, or push AgentSec while its license decision blocks public release.
The compatibility database still consumed by the guide commands is
examples/commands/resources/threat-db.yaml.
test cases passed, vs 6 / 20 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.