The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from Jira Cloud, such as issue summaries, descriptions, and comments, which can be modified by external or untrusted users. This data is then used within the agent's context, creating a potential for indirect prompt injection. The skill also possesses write capabilities (creating/updating issues and comments) which could be triggered by injected instructions.
Ingestion points: scripts/jira-api.py is used to fetch issue details (GET /issue/{key}) and search results (GET /search).
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the documentation.
Capability inventory: The skill allows creating issues, updating fields, adding comments, and transitioning statuses via POST and PUT methods.
Sanitization: The helper script does not sanitize the content retrieved from or sent to the Jira API, acting as a direct transport layer.
[COMMAND_EXECUTION]: The skill operates by executing the provided scripts/jira-api.py script. The script uses standard Python libraries (urllib, json, base64) and includes safe practices such as loading configuration from a .env file and enforcing HTTPS for API connections.
[CREDENTIALS_SAFE]: The skill correctly handles sensitive Jira API tokens by instructing the user to use environment variables. It does not contain any hardcoded credentials, only placeholders for configuration.

jira