The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill provides instructions for a workflow that uses python3 -c to decode URL-encoded segments of asta:// URIs. The $ENCODED_INDEX_URL variable is interpolated directly into the Python command string using single quotes. If a maliciously crafted URI containing a single quote is processed, it could escape the string and execute arbitrary Python code.
[COMMAND_EXECUTION]: The skill uses shell commands like curl and the asta CLI tool to handle data from external sources. The construction of these commands using variables derived from remote URLs creates a risk of command injection if the input strings are not strictly validated.
[INDIRECT_PROMPT_INJECTION]: The skill indexes documents from various external sources including web URLs and cloud storage. It extracts metadata like names and summaries which can then be processed by the agent.
Ingestion points: Remote documents fetched via asta documents fetch or added via asta documents add, and remote YAML index files downloaded to /tmp/.
Boundary markers: Absent. The skill does not define delimiters or provide instructions to ignore embedded prompts in document content.
Capability inventory: The skill has broad capabilities including shell execution (Bash), file system access (Write to .asta/ and /tmp/), and network operations via curl, aws, and gcloud.
Sanitization: Absent. The skill does not describe any validation or sanitization for the content of processed documents or index files.
[EXTERNAL_DOWNLOADS]: The skill guides the agent to install a CLI tool from the author's official GitHub repository. This download is documented as part of the intended functionality for index management.

asta-documents