The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill contains instructions to extract a GitHub authentication token from the sensitive configuration file ~/tau.yaml using awk for use in local API requests.
[COMMAND_EXECUTION]: The skill utilizes several shell commands to interact with the local environment and network, including curl, awk, and the vendor-specific CLI tools tau and dream.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by ingesting and storing external data in the form of build logs.
Ingestion points: Build logs are fetched and stored locally via the tau query logs command as described in Step 5.
Boundary markers: None identified; the skill does not specify delimiters or warnings for the agent when processing these logs.
Capability inventory: The skill uses curl for network requests, awk for file reading, and tau for file-write operations.
Sanitization: There is no evidence of validation or sanitization of the log content before it is processed or presented to the user.

diagnosing-dream-builds