The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via untrusted data sources. * Ingestion points: Incident signals are fetched using the webfetch tool and provided via 'incident summaries' inputs. * Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the ingested data. * Capability inventory: The skill uses github.create_issue to write content to external repositories. * Sanitization: The instructions do not specify any sanitization or filtering of the fetched content before it is processed by the issue-writer agent.
[DATA_EXFILTRATION]: The skill implements a workflow that reads data from external sources and writes it to a GitHub repository. While this is the stated purpose, it creates a potential channel for sensitive data exposure if the input signals contain secrets or if the output repository is not properly secured or is controlled by an attacker.

github-issue-reporter