The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill includes recipes for mutual TLS (mTLS) configuration that require accessing sensitive local files, such as client certificates and private keys, using the --client-cert and --client-key flags.
[COMMAND_EXECUTION]: The skill provides templates to run the telemetrygen utility, which is capable of generating significant network traffic and could be used for load testing or potential service disruption if misconfigured.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it interpolates untrusted user input directly into command-line arguments. 1. Ingestion points: User-provided telemetry attributes, service names, and log bodies (SKILL.md). 2. Boundary markers: Absent; the instructions do not use delimiters or provide warnings to ignore embedded instructions in the data. 3. Capability inventory: Execution of the telemetrygen binary via the shell (SKILL.md). 4. Sanitization: Absent; the skill does not specify any validation or escaping of user-provided strings before they are used in commands.
[EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install telemetrygen from the official OpenTelemetry project repositories on GitHub and the GitHub Container Registry.

telemetrygen