Security audit
Security checks for vulnerabilities and agentic risk
This skill only guides job-fit evaluation and proposal drafting from a profile and template, with no hidden execution or persistence.
Install this if you want an assistant to evaluate job posts against Caleb's packaged profile and draft short proposals. Be aware that in a mixed-purpose chat it may treat unrelated messages as job listings, so use it in a dedicated job-posting context or invoke it explicitly.
The catch-all trigger phrase 'or similar' makes the activation boundary ambiguous, which can cause the skill to run on requests that are only loosely related to job evaluation or proposal writing. This broadens the attack surface for accidental or adversarial invocation, though the impact is limited by the skill's relatively narrow downstream actions.
The instruction to treat every incoming group message as a job listing removes basic input validation and can cause the skill to activate on unrelated or maliciously crafted messages. In practice, this can lead to incorrect processing, unintended use of profile/reference data, and proposal generation from non-job content, increasing the chance of prompt confusion or data misuse.
No suspicious patterns detected.
