The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data retrieved from changelogs and external websites.
Ingestion points: Data enters the agent context through the releases admin source fetch command, the search tool (which processes changelog_chunk hits), and the WebFetch tool used to fill data gaps from external URLs.
Boundary markers: The instructions do not specify the use of delimiters or specific instructions to the model to ignore embedded commands within the fetched release data or changelogs.
Capability inventory: The skill uses the releases CLI suite and potentially writes analysis results to files, though it does not demonstrate arbitrary shell execution or sensitive system access.
Sanitization: There is no evidence of sanitization or filtering applied to the fetched content before it is processed for summarization or comparison.
[COMMAND_EXECUTION]: The skill relies on a suite of releases CLI commands (including list, fetch, tail, search, summary, and compare) to interact with its data sources. These tools appear to be specialized utilities for the skill's primary purpose.

analyzing-releases